# MemoryLane: full content > Every page of trymemorylane.com as markdown, concatenated for LLM ingestion. ================================================================================ # MemoryLane (Home) Source: https://trymemorylane.com/ ================================================================================ --- title: MemoryLane | Map Your Processes, Drive AI Adoption description: "Map manual workflows in your company and see how to redesign and automate them." url: https://trymemorylane.com --- # Map your processes, drive AI adoption Map manual workflows in your company and see how to redesign and automate them. ## Your team is ready for AI. You just need to know where to start. - **AI tools approved. Now what?** You've rolled out Claude, Copilot, or automation platforms, but your team is still copy-pasting between spreadsheets because nobody mapped which workflows to replace first. - **Audits miss the real work.** Process interviews capture what people remember. They miss the invoice downloads, the confirmation emails, the copy-paste that happens a hundred times a week. - **AI adoption without guesswork.** MemoryLane shows you exactly where your team's time goes and which workflows are worth automating first. ## Your AI adoption roadmap. Built by AI. Most consultants hand you a slide deck after weeks of interviews. MemoryLane analyzes your team's real workflows and delivers a prioritized roadmap: which processes to transition, which tools to use, and the ROI for each. Example roadmap rows: | Workflow | Hours / yr | Annual cost | Action | | ---------------------- | ------------ | ----------- | ---------------- | | Status update sync | 340 hrs / yr | $28,900 | Transition to AI | | Weekly report assembly | 280 hrs / yr | $23,800 | Transition to AI | | Request routing | 230 hrs / yr | $19,550 | Transition to AI | Download a sample roadmap: https://trymemorylane.com/memorylane-case-study.pdf ## Four ways teams put MemoryLane to work - **Map Processes.** See exactly how a workflow really happens, no interviews needed. - **Create SOPs.** Turn a captured workflow into a step-by-step SOP your team can follow. - **Onboard New Employees.** New hires learn how the team actually works from captured workflows, not a stale wiki page. - **Discover Automatable Tasks.** Try it yourself first. Run MemoryLane on your own machine and see which of your daily tasks are worth handing to an AI agent. ## Pricing Start with yourself. Expand to your team when you're ready. ### Individual ($100/mo) Try it yourself first. Run MemoryLane on your own machine and see what it discovers: a lightweight preview of what you get. - MemoryLane maps your daily workflows in the background - Generate your AI adoption snapshot with Claude in one click - See which tasks are eating the most time - See which of your workflows AI can handle today - Captured locally; inference via ZDR-compliant endpoints - Email support ### Enterprise (Custom) The full engagement: we deploy across your team and deliver your AI adoption roadmap with implementation guidance. - Everything in Individual - Workflow analysis across your entire team - We help you execute the transition, not just plan it - Delivered roadmap with tool recommendations and ROI - Fast team-wide deployment (on-premises available) - Privacy controls and data policies built in ZDR (Zero Data Retention): inference requests are routed through OpenRouter to ZDR-compliant provider endpoints; prompts and responses are not stored or used for training. See the [privacy policy](https://trymemorylane.com/privacy). ## From kickoff to roadmap No workflow mapping, no months of consulting. 1. **Align.** We meet with your leadership to define goals, scope the engagement, and agree on what success looks like: genuine ROI, not hype. 2. **Deploy.** We install MemoryLane on your team's devices. It runs in the background, analyzing workflows and measuring time across applications; recordings stay on each device. 3. **Deliver.** We help you analyze the data and produce an AI adoption roadmap: which workflows to transition, which tools fit best, the ROI for each, and a clear order of operations. ## Private by design Your data, your terms. - **No screenshots stored.** Each screenshot becomes a short text summary, then gets deleted. Never stored, never used to train AI models. - **Choose your AI.** Our API key (ready in under a minute, Zero Data Retention per provider contract), your own OpenRouter key, or local AI that runs fully offline for free. - **Data stored locally.** Summaries live on your device, not on MemoryLane servers. Only you can open them. Who sees my data? Nobody at MemoryLane. Your screen data never touches our servers. Screenshots pass through the AI provider you chose for a few seconds of processing under Zero Data Retention terms, then they're deleted. Full details: https://trymemorylane.com/security ## FAQ ### What does your tool do? It runs in the background and captures how work really happens, then maps it out for you to see. No interviews or workshops. Its goal is to surface repetitive tasks worth automating and help you turn them into process maps, SOPs, and ready-to-use AI agent prompts. ### How is it different from n8n and Zapier? n8n and Zapier are automation tools you use once you know exactly what automations to build. We come a step before that. MemoryLane spots the repetitive tasks you do and tells you which ones you can automate and how. This way you don't have to figure it out yourself. ### How exactly does the technology work? MemoryLane runs in the background and takes periodic screenshots of your screen. Each screenshot is sent to a Zero Data Retention (ZDR) AI endpoint, which analyzes it and returns a short text summary of the activity. That summary is what gets saved, and it's saved locally on your laptop. The screenshot itself is deleted right after the AI provider processes it: it's never stored and never used to train AI models, per contractual commitments from providers like Google and Anthropic. MemoryLane itself never sees the screenshots or has access to your data, since everything stays local on your device. ### Can I exclude apps or websites from being captured? Yes. Anyone can blacklist any desktop application or website URL, and MemoryLane will never capture it: useful for personal banking, private messages, or anything not related to work. With a Central rollout, your organization can also set the blacklist centrally, so it applies to everyone without individual setup. ### Is my data secure? Yes. Screenshots are sent to Zero Data Retention (ZDR) AI endpoints, which means no screenshots are ever stored anywhere or used for training. The resulting textual data is saved on each user's device by default and shared centrally with their explicit consent. Organizations that need full data isolation can deploy MemoryLane on-premise. [Read more on our security page](https://trymemorylane.com/security). ### How does this work for me vs. my whole team? We recommend you start with the Individual plan ($100/mo) first. It captures your own work and shows you what's worth automating within days. When you're ready to go further, the Central rollout deploys the same capture across your team, maps workflows org-wide in a central place, and delivers an AI adoption roadmap with an implementation help. ### How long before I see results? Most people see their first automation suggestions within 7 days. The longer it runs, the more patterns it finds. ### How does pricing work? The Individual plan is $100/month so you can try it on your own workflows first. Central rollout is scoped per organization. Book a call and we'll give you a quote. ### How disruptive is the process? Minimal. Our software runs in the background and your team keeps working normally. ### What is task mining? Task mining observes how work actually happens on people's screens and finds patterns automatically. Instead of interviewing people about their workflows, the software watches app switches, repeated steps, and time spent, then ranks processes by automation potential with KPI data like hours per week and estimated cost. It is process mapping without the manual work. [Learn more about task mining for small teams](https://trymemorylane.com/blog/task-mining-for-small-teams). ### What is process mapping? Process mapping means documenting how work gets done, step by step. Most teams use business process mapping software like Visio, Miro, or Lucidchart to draw flowcharts manually. The problem is that manual mapping depends on interviews, and people cannot accurately describe work they do on autopilot. MemoryLane automates this by observing real screen activity and producing ranked process maps you can export as BPMN, Mermaid, PDF, or PNG. [Compare process mapping tools](https://trymemorylane.com/blog/best-process-mapping-software). ## Links - [User Guide](https://trymemorylane.com/guide) - [Become a Partner](https://trymemorylane.com/partners) - [Privacy Policy](https://trymemorylane.com/privacy) - [Security & Technical Overview](https://trymemorylane.com/security) - [Terms and Conditions](https://trymemorylane.com/terms) - [GitHub](https://github.com/deusXmachina-dev/memorylane) ================================================================================ # MemoryLane | Become a Partner Source: https://trymemorylane.com/partners ================================================================================ --- title: MemoryLane | Become a Partner description: "Partner with MemoryLane to help your clients map workflows and build AI adoption roadmaps backed by real operational data." url: https://trymemorylane.com/partners --- # Partner with MemoryLane Help your clients map their workflows and build AI adoption roadmaps backed by real operational data, not guesswork. ## Who we partner with - **Consulting firms.** Add workflow intelligence to your AI strategy engagements. Deliver roadmaps backed by real data instead of interviews. - **System integrators.** Use MemoryLane to scope automation projects accurately. Know which processes to target before you build. - **PE Rollups.** Map real workflows across a portfolio company in weeks, then hand over a prioritized automation roadmap with hours and dollars attached. A repeatable value creation play for every new acquisition. - **Managed service providers.** Offer AI readiness assessments as a service. Deploy MemoryLane across client teams and deliver prioritized roadmaps. ## What partners gain - **Data-backed AI strategy.** Replace guesswork with workflow data. Your recommendations are grounded in how teams actually work. - **Privacy-first architecture.** Data stays on client devices. No screenshots stored, no raw data shared. Your clients trust the process. - **Faster engagements.** Skip weeks of process interviews. MemoryLane maps workflows in the background while your team focuses on strategy. - **Flexible models.** Referral, reseller, or embedded. We fit your business model, not the other way around. ## Get started Tell us about your practice and we will explore how MemoryLane fits. No commitment, no pitch deck required. ================================================================================ # User Guide Source: https://trymemorylane.com/guide ================================================================================ --- title: User Guide | MemoryLane description: Set up MemoryLane in 3 steps. About 5 minutes. url: https://trymemorylane.com/guide --- # Get Started with MemoryLane Set up MemoryLane in 3 steps. About 5 minutes. ## Step 1: Install ### Mac 1. Open the downloaded `.dmg` file (Individual) or `.pkg` file (Team). 2. Drag **MemoryLane** into your **Applications** folder (Individual) or follow the installer prompts (Team). 3. Open MemoryLane from Applications. 4. Grant **Screen Recording** and **Accessibility** permissions when prompted (System Settings > Privacy & Security). You may need to restart the app. Requires a Mac from 2020 or newer (Apple Silicon M1 chip or later). ### Windows 1. Double-click the downloaded `.exe` file (Individual) or `.msi` file (Team) and follow the prompts. 2. Open MemoryLane from the Start Menu or desktop shortcut. ### Security warnings This is normal. MemoryLane needs screen recording access to work. Your system is doing its job well by letting you know and requesting an approval. - **Mac:** If macOS says the app "can't be opened," right-click (or Control-click) the app and select **Open**. This only happens the first time. - **Windows:** You may see a blue popup saying "Windows protected your PC." This is a standard check for newly released apps. Click **More info**, then **Run anyway**. ### Team activation 1. After granting permissions, you'll see the Activate Device screen. 2. Paste the activation code from your email (starts with `tt_`). 3. Review and accept the consent document. 4. Click **Activate Device**. ## Step 2: Set privacy rules Apps and websites on your blacklist are excluded from capture. 1. Open **Settings** (gear icon, top right). 2. Go to **Exclude Apps** to block applications. 3. Go to **Exclude Websites** to block websites. Search for the app or website, flip the toggle. Changes apply immediately. ## Step 3: Start capture Press the green **Start Capture** button. --- ## Common Questions ### What does MemoryLane observe? - **Screenshots** at regular intervals (e.g. a snapshot of your Excel spreadsheet) - **On-screen text (OCR)** (e.g. "Q2 Budget Draft", "Sheet1 - Revenue", "=SUM(B2:B50)") - **Active window titles and app names** (e.g. "Microsoft Excel - Q2 Budget.xlsx") - **Time spent in each application** (e.g. 45 min Excel, 20 min Gmail, 10 min Slack) - **Tab and window switching patterns** (e.g. switched Excel and Gmail 12 times) ### How do I exclude an app or website from being captured? Open Settings (gear icon), go to **Exclude Apps** or **Exclude Websites**, search for the app, and flip the toggle. You can also use **Auto-fill from activity**: click Auto-fill, use the apps you want blocked for 2 minutes, then toggle the ones detected. ### What happens when a privacy rule matches? When you switch to a blocked app or website, capture pauses automatically and resumes when you leave. Privacy rules only match the window you're currently using. If a blocked app is visible in the background, it can still appear in screenshots. ### How do I pause or stop capture? Press **Stop Capture** in the app. MemoryLane won't restart capture on its own, so if you leave it off, no data will be collected the next day. Remember to turn it back on after a temporary pause. ### Where is my data stored? Only short text summaries and on-screen text (OCR) are saved locally on your computer. Screenshots are processed and immediately deleted. Nothing is stored on our servers. You can export your data: go to **Settings > Data** and click **Export Database (.zip)**. The export stays on your computer and won't be shared with anyone unless you send it yourself. **Team:** In the app's settings, you choose whether anything is shared with your company's admin dashboard. Options: Off (default, nothing leaves your device), Summary (activity summaries only), or Detailed (summaries and on-screen text). ### The app is frozen or won't open If MemoryLane is unresponsive, not opening when you click it, or showing a blank screen, restart it. **Quick restart (Mac):** Open the MemoryLane window and press **Cmd + R** to reload. **Full restart (Mac):** 1. Right-click the MemoryLane icon in the menu bar (top right of your screen) and click **Quit**. 2. If nothing happens, open **Activity Monitor** (search for it in Spotlight with Cmd + Space). 3. Search for **MemoryLane** in the list. 4. Select it, click the **X** button in the toolbar, then click **Force Quit**. 5. Open MemoryLane again from your Applications folder. **Full restart (Windows):** 1. Right-click the MemoryLane icon in the system tray (bottom right of your screen, near the clock) and click **Quit**. 2. If nothing happens, press **Ctrl + Shift + Esc** to open **Task Manager**. 3. Find **MemoryLane** in the list. 4. Right-click it and click **End task**. 5. Open MemoryLane again from the Start Menu or desktop shortcut. If the app still won't start after a full restart, try restarting your computer. ### How do I update the app? **Individual:** Updates are automatic. When one is ready, you get a notification. Click it to restart and install. Missed it? Open the menu bar (Mac) or system tray (Windows) menu and click **Install Update Now (Restart)**. **Team:** Download the latest installer and run it again. Activation persists, no re-activation needed. ### How do I check my version? The version number appears in the menu bar (Mac) or system tray (Windows) menu as a **MemoryLane v0.xx.x** row. --- ## Analyzing Your Data You have two options: ### 1. In-app suggestions Built in. No setup needed. ### 2. AI-powered deep analysis (recommended) Reports, time audits, automation tips via Claude. Requires a [paid Claude account](https://claude.ai). One-time setup. #### Setup 1. **Connect Claude:** Open Settings > Connections & Data > Integrations, then connect Claude Cowork and/or Claude Code. No extra software needed. 2. **Add data analysis skills:** Skills teach Claude how to read your MemoryLane data like a professional data analyst. Copy and paste `deusXmachina-dev/memorylane` into the Marketplace plugin field. #### Available skills | Skill | Description | | ------------------------- | --------------------------------------------------------------------------------------------------------- | | `last-30-minutes` | Summarizes what you've been doing in the last 30 minutes. Quick recap, no analysis. | | `discover-patterns` | Lightweight process analysis. Finds repeated workflows and generates an HTML report. | | `pattern-to-pdf` | Lightweight PDF report from a detected pattern: process map, steps, frequency. | | `process-analyst-new` | Deep process analysis. Breaks down every repeated task with time and cost estimates. Most powerful skill. | | `report-new` | Turns process-analyst-new results into a shareable PDF report. | | `automation-instructions` | Step-by-step build instructions for automating a workflow in the tools and constraints you specify. | #### Quick analysis example First, try: ``` /discover-patterns in my data. I am a product operations executive in Acme, a pet food e-commerce store in the UK. My main responsibilities are to oversee our supply ordering workflows using Shopify, Netsuite ERP and Gmail, and customer support. ``` Then run: ``` /automation-instructions for n8n where possible and Claude Cowork where human involvement is needed. ``` #### Deep analysis example First, try: ``` /process-analyst-new I am a product operations executive in Acme, a pet food e-commerce store in the UK. My main responsibility is to oversee our supply ordering workflows using Shopify, Netsuite ERP and Gmail. The process starts by checking the status of our supplies in Shopify for each vendor on the Supplies page and ends when I hit Order for each supplier in Netsuite. ``` Then run: ``` /report-new build a report outlining the identified automation recommendations so I can present it at our next weekly standup. ``` Or: ``` /automation-instructions for n8n where possible and Claude Cowork where human involvement is needed. ``` ### Claude stopped working? If Claude or Cursor updates, the connection can break. Open Settings > Connections & Data > Integrations. If you see a **Reconnect** button, click it. --- Don't have time to analyze data? We can help. [Book a call](https://calendar.app.google/2rJgu2Ah3kWaMApG8). Email us at founders@trymemorylane.com ================================================================================ # Changelog Source: https://trymemorylane.com/changelog ================================================================================ # Changelog What's new in the MemoryLane app, newest first. Short notes in plain English. ## 2026-08-04 v1.5.5 - **Patterns front and center** Patterns is now the home screen, showing hands-on time, estimated monthly hours, and a 4-week trend for each one. - **Turn a pattern into an AI agent** Every pattern now comes with a step-by-step recipe with personal details stripped out. Review it, then copy it or open it in Claude to automate the task. - **Company devices stay up to date** IT-managed Macs and Windows machines now get updates automatically. ## 2026-07-10 v1.4.0 - **Find tasks you keep repeating** MemoryLane looks back through up to 60 days of your work to surface repeated tasks. The analysis runs on your device. - **Smarter AI models, automatically** The app picks the best available models on startup, so you get improvements without updating. - **Cleaner look** Updated colors and tighter spacing across the app. ## 2026-06-28 v1.3.0 - **Company-wide privacy blocklist** IT can now block chosen apps from capture for the whole company, and the list updates on every computer within minutes. - **More dependable syncing** Data now syncs to your company workspace exactly once per day, even if the app restarts. ## 2026-06-24 v1.2.1 - **Smoother Mac setup** MemoryLane now shows up correctly in the macOS Screen Recording permission list during setup. ## 2026-06-23 v1.2.0 - **Pause from the menu bar** Pause and resume capture in one click. Capture also resumes safely after your computer wakes from sleep. - **Reading time counts** Time spent reading, not just clicking and typing, is now captured, so quiet work shows up in your day. - **Smarter pattern detection** We rebuilt how repeated tasks are recognized, so the patterns you see are more accurate. - **Faster syncing** Uploads are compressed and no longer slow the app down. - **One-click log export** Send us your app logs straight from Settings when something goes wrong. ## 2026-06-04 v1.1.0 - **Bring your own AI models** Point MemoryLane at your own model server, including ones on your private network, with a new step-by-step guide. - **Easier updates** When a new version is ready, a banner in the app lets you relaunch and update in one click. ## 2026-06-02 v1.0.0 - **Activities and Patterns, redesigned** Every detected pattern now shows the real moments behind it, so you can check the evidence yourself. - **New setup experience** A rebuilt first-run flow, a resizable main window, and plenty of Windows polish. - **Move your data** Import a database export to carry your history to a new computer. ## 2026-05-10 v0.26 - **Simpler company setup** Your activation code now carries everything needed to connect to your company workspace, and devices already approved by IT skip the consent screen. ## 2026-04-29 v0.25 - **One-code activation** A single activation code sets up company devices, with a clear consent step before any capture starts. ## 2026-04-19 v0.23 - **Claude connects directly** Connect Claude to your MemoryLane data without installing anything extra. - **Better privacy controls** Searchable lists let you exclude any app or website from capture, with one-click suggestions based on what you use. - **Enterprise Mac installer** Company rollouts on Mac now use a standard notarized installer, and Windows upgrades got more reliable. ## 2026-04-16 v0.22 - **You choose what syncs** Pick how much detail leaves your device: nothing, summaries only, or full detail. ## 2026-04-14 v0.21 - **Sensitive data stays local** Sensitive information is stripped from the database before anything is uploaded to your company workspace. ## 2026-04-12 v0.20 - **Automatic company sync** Enterprise devices now sync on a schedule, plus a manual sync button when you don't want to wait. ================================================================================ # Security & Technical Overview Source: https://trymemorylane.com/security ================================================================================ # Security & Technical Overview ## Core Principle **You only share what you choose to share.** Your data stays on your device. Screenshots are captured, processed locally or via secure AI, and then cleaned up after processing. A background sweep every 10 minutes removes screenshot files older than 1 hour. The semantic insights extracted are stored only on your machine. Nothing leaves your device unless you explicitly choose to share it, with the following exceptions: (1) if using Scenario A, the app contacts `api.trymemorylane.com` on startup to validate your subscription using a device identifier; (2) on first use, the local embedding model is downloaded from `huggingface.co`. ## Basics ### Screenshot Lifecycle 1. **Capture**: Screenshots are taken at approximately 1-second intervals during active activity sessions. Sessions are bounded by application switches, 5-second inactivity gaps, and a 5-minute maximum duration. 2. **Prepare**: Depending on the capture mode: - **Auto mode** (default): Attempts video first and falls back to image mode if video processing is unavailable or fails. - **Video mode**: Captured frames are stitched into a 1 FPS video via FFmpeg. The video is sent alongside activity metadata for AI summarization. - **Image mode**: Up to 6 key screenshots are sampled based on visual difference and interaction timing, then fed as images into a vision model. 3. **Process**: AI extracts semantic meaning (what you were doing, context). Local OCR (macOS Vision framework / Windows native OCR) also extracts on-screen text. Local embeddings are generated for semantic search using the `@huggingface/transformers` library with the all-MiniLM-L6-v2 model. On first use, the embedding model (~100 MB) is downloaded from `huggingface.co` and cached locally for subsequent use. 4. **Store**: Semantic data, OCR text, and embeddings are saved to a local SQLite database on your device. 5. **Clean Up**: Screenshot and video files are cleaned up after processing. A background sweep runs every 10 minutes and removes any screenshot or video files older than 1 hour. ### Local Storage - **macOS**: `~/Library/Application Support/MemoryLane/` - **Windows**: `%APPDATA%\MemoryLane\` - **Logs (macOS)**: `~/Library/Logs/MemoryLane/` - **Logs (Windows)**: `%APPDATA%\MemoryLane\logs\` - **Format**: SQLite database - **Ownership**: You control your data completely ## AI MemoryLane offers three deployment scenarios for AI processing: ### Scenario A: OpenRouter (Default) **How it works:** - Data goes **directly from your device to OpenRouter**: SenseFlow does not proxy or route your data - OpenRouter load-balances and geo-balances across AI providers: - **Video mode**: A stitched 1 FPS video is sent alongside activity metadata to AI models (e.g., Gemini 2.5 Flash Lite, Gemini 3 Flash, Molmo 2 8B) via ZDR-compliant endpoints - **Image mode**: Up to 6 sampled screenshots are fed into AI models (e.g., Mistral Small 3.2, Gemini 2.5 Flash Lite) via ZDR-compliant endpoints - **Pattern detection (Pro)**: Kimi K2.5 via ZDR-compliant endpoint. Receives activity metadata including text summaries, window titles, app names, domain names, and timestamps (no images or screenshots) - Zero Data Retention (ZDR) is enabled at the OpenRouter account level for managed API keys (see ZDR section below) - Only semantic data (text describing what was on screen) returns to your device - Screenshots are cleaned up after processing (background sweep removes files >1 hour old) - Local OCR and embeddings are computed entirely on-device ### Scenario B: Your Cloud AI Endpoint **How it works:** - Configure MemoryLane to use your own AI endpoint - You control the AI provider and data handling - MemoryLane never sends your data to us - Self-setup or paid monthly support available **Recommended providers:** - **Image/video processing**: Google Vertex AI (Gemini models). These are the only models that have consistently produced reliable results for screenshot and video analysis. Other vision models may work but have not been validated to the same standard. - **Text analysis and pattern detection**: Anthropic (Claude models). Recommended for activity summarization, pattern detection, and report generation. **Our recommendation:** Most teams are better off with Scenario A. You get zero data retention by default, no infrastructure to manage, and no provider agreements to negotiate. Scenario B takes several days of engineering work, adds roughly 30% to project costs, and if you need ZDR with Google Vertex AI, that is a separate agreement you have to negotiate with Google. If you still want to go this route, email founders@trymemorylane.com and we will help you scope it. ### Scenario C: Fully Local Models **How it works:** - AI model runs entirely on your device - **No data ever leaves your machine** - Complete privacy and air-gap capability - Self-setup or paid monthly support available **Our recommendation:** This is not practical for most teams. Running local AI models at scale needs hardware closer to a small data center than a laptop, costing $50,000+. A small pilot (up to 3 users) is doable but still needs dedicated engineering. Email founders@trymemorylane.com if you want to talk through what is realistic for your setup. ## Technical ### What Installs | Component | Description | | ------------------------- | ---------------------------------------------------------------------------------------------------- | | MemoryLane App | Desktop application (macOS Apple Silicon, Windows x64) | | Local SQLite DB | Stores semantic data, OCR text, and embeddings (unencrypted; protected by OS-level file permissions) | | FFmpeg | Bundled for video stitching of screenshot sequences | | @huggingface/transformers | Local embedding generation (all-MiniLM-L6-v2 model, downloaded from huggingface.co on first use) | | Native sidecar binaries | Swift (macOS) and Rust (Windows) binaries for active window detection and browser URL extraction | | Config files | User preferences, AI endpoint settings | ### What It Captures - **Screenshots** at ~1-second intervals during activity sessions (bounded by app switches, 5-second inactivity gaps, and 5-minute maximum duration) - **Interaction metadata** (used ephemerally for session windowing, not stored in the database): keyboard session durations and key counts, mouse click positions (x, y coordinates), scroll events. No keystroke content is recorded - **Window context**: active window titles, process names, bundle identifiers, and browser URLs (for categorization). Browser URLs are extracted via the macOS Accessibility API or equivalent for supported browsers (Chrome, Safari, Edge, Brave, Arc, Vivaldi, Opera) - **OCR text**: on-screen text extracted locally via macOS Vision framework or Windows native OCR - **Video clips**: screenshot sequences stitched locally via FFmpeg - **Embeddings**: semantic vectors generated locally via @huggingface/transformers (all-MiniLM-L6-v2) - User can pause capture anytime directly in the app - Screenshots are captured from all visible windows, including incognito/private browser windows, when they are the active window during a capture interval ### What Gets Transmitted | Scenario | What's Sent | Where | | -------------------------- | ------------------------------------------------------------------------------------- | ------------------------------------------------------- | | A (OpenRouter, Video mode) | Stitched 1 FPS MP4 video + activity metadata | Direct to OpenRouter → ZDR-compliant provider endpoints | | A (OpenRouter, Image mode) | Up to 6 compressed screenshots | Direct to OpenRouter → ZDR-compliant provider endpoints | | A (Pattern Detection) | Activity metadata: text summaries, window titles, app names, domain names, timestamps | Direct to OpenRouter → ZDR-compliant provider endpoints | | A (Subscription/Auth) | Device identifier | `api.trymemorylane.com` (automatic on startup) | | A (Payments) | Payment info | Stripe (via browser redirect) | | First run (all scenarios) | Model download request | `huggingface.co` (embedding model, ~100 MB, one-time) | | B (Your Endpoint) | Per your config | Your infrastructure | | C (Local) | Nothing | Stays on device | ### Network Requirements (Scenario A) | Endpoint | Port | Purpose | | ----------------------- | ----------- | ---------------------------------------------------------------------------- | | `api.openrouter.ai` | 443 (HTTPS) | AI processing | | `api.trymemorylane.com` | 443 (HTTPS) | Subscription validation, licensing (device ID sent automatically on startup) | | `github.com` | 443 (HTTPS) | Application updates | | `huggingface.co` | 443 (HTTPS) | One-time embedding model download on first use | ### System Requirements | Platform | Minimum | | -------- | ------------------------------------------ | | macOS | Apple Silicon (ARM64), 4GB RAM, 500MB disk | | Windows | Windows 10+ (x64), 4GB RAM, 500MB disk | ## Security ### Encryption | Context | Method | | ------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | In transit | HTTPS/TLS for all API communications (TLS version negotiated by the OS and server) | | Credentials at rest | API keys and device ID encrypted via Electron safeStorage (uses macOS Keychain / Windows DPAPI to derive encryption keys; encrypted data stored in local config files) | | Local database | SQLite (unencrypted). protected by OS-level file permissions and user account access controls. Any process running as the same OS user can access the database. | ### Zero Data Retention (Scenario A) When using Scenario A, screenshots and video are sent **directly from your device to OpenRouter** (SenseFlow never sees or proxies this data). OpenRouter load-balances and geo-balances requests across AI providers. **Managed API keys**: SenseFlow provisions OpenRouter API keys with Zero Data Retention (ZDR) enabled at the account level. When ZDR is enabled, OpenRouter only routes requests to endpoints it classifies as ZDR-compliant. Requests that cannot be routed to a ZDR-compliant endpoint will fail rather than fall back to a non-ZDR provider. ### Bring Your Own Key (BYOK) If you use your own OpenRouter API key, ZDR depends on your own OpenRouter account settings. You are responsible for enabling ZDR in your OpenRouter account if desired. OpenRouter's ZDR is a routing-level control, determining which provider endpoints your requests can be sent to. Individual AI providers may have their own data retention policies independent of OpenRouter. Customers with specific compliance requirements should review provider policies directly. ### Interaction Monitoring MemoryLane monitors user interaction to detect activity sessions and provide productivity context. Here is what is and is not recorded: | Input Type | What IS Recorded | What is NOT Recorded | | ------------ | ------------------------------------------------------------------- | ---------------------------------------------------------- | | Keyboard | Session duration, key count, active window title | Keystroke content (what you type) | | Mouse | Click position (x, y coordinates) | Click count (not persisted), what was clicked on (content) | | Scroll | Cumulative scroll amount, direction (vertical/horizontal) | Scroll content | | Window focus | Window title, process name, bundle ID, browser URL, previous window | Window content (captured separately via screenshots) | Interaction data is collected via `uiohook-napi` (keyboard, mouse, scroll) and native sidecar binaries (window/app changes, browser URLs). Interaction metadata is used ephemerally for session windowing and is not durably stored in the database. It is never transmitted to SenseFlow or any third party. ### MCP Server Integration MemoryLane includes an optional Model Context Protocol (MCP) server that allows compatible AI assistants (e.g., Claude, Cursor) to query your local activity data. - **Protocol**: Local stdio-based MCP server. On first use, the embedding model is downloaded from `huggingface.co` if not already cached; all subsequent communication is local only - **Configuration**: User must explicitly configure their AI assistant to connect to the MCP server - **Data scope**: AI assistants can query semantic data, activity summaries, raw OCR text, and search your local history - **No data sent to SenseFlow**: The MCP server operates entirely locally. SenseFlow has no access to MCP queries or responses - **User control**: The MCP server can be disabled at any time in app settings ### Pattern Detection (Pro Feature) MemoryLane Pro includes pattern detection that identifies recurring work patterns and provides productivity insights. - **Input**: Activity metadata including text summaries, window titles, app names, domain names, and timestamps for the day's sessions (no images or screenshots) - **Processing**: Activity metadata is sent to Kimi K2.5 via OpenRouter once per day (routed to a ZDR-compliant endpoint) - **Output**: Text-based pattern descriptions and productivity insights - **Storage**: Pattern data stored locally in the SQLite database - **No images transmitted**: Pattern detection receives only text, never screenshots or video ### PII & Sensitive Data Protection MemoryLane protects personally identifiable information through multiple layers: **Architecture-level protection.** AI processes screenshots into short text summaries describing what you were doing. Original screenshots and video are deleted after processing (background sweep removes files older than 1 hour). The data stored long-term is semantic descriptions, not raw screen captures. **Application and website blocklists.** Users (or IT admins) can blocklist specific applications and websites from capture entirely. Blocked apps are never screenshotted. **Regex-based PII redaction.** Before summaries are stored, MemoryLane runs pattern-matching to detect and redact common PII formats: | Category | Examples | | ---------------------------- | -------------------------------------------------------------------- | | Email addresses | name@example.com | | Phone numbers | US, AU, NZ, PH, MY, CA, DE, and international formats | | Government / tax identifiers | SSN (US), TFN (AU), IRD (NZ), TIN (PH, MY), SIN (CA), Steuer-ID (DE) | | Financial identifiers | Credit card numbers, IBANs | Pattern coverage is regularly expanded. If your region or identifier type is not listed, contact founders@trymemorylane.com. ### Supported AI Models (Scenario A) | Capture Mode | Models Used | Input | | ----------------------- | ------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------- | | Auto mode (default) | Attempts video models first, falls back to image models | See below | | Video mode | Gemini 2.5 Flash Lite (preview), Gemini 3 Flash (preview), Molmo 2 8B, via OpenRouter ZDR-compliant endpoints | 1 FPS stitched video + activity metadata | | Image mode | Mistral Small 3.2, Gemini 2.5 Flash Lite, via OpenRouter ZDR-compliant endpoints | Up to 6 sampled screenshots (as images) | | Pattern detection (Pro) | Kimi K2.5, via OpenRouter ZDR-compliant endpoints | Activity metadata: summaries, window titles, app names, domains, timestamps | All models are accessed via OpenRouter API. The specific models used may be updated over time; current defaults are listed above. ### User Controls | Control | Description | | ----------- | --------------------------------------------------------------------------------------------------- | | Pause | Stop all capture with one click | | View data | See all semantic data stored locally | | Export data | Export the local database as a ZIP file containing the SQLite database | | Uninstall | Remove the application. User data directories must be manually deleted (see Uninstallation section) | ### Audit Trail - Local logs available at: - macOS: `~/Library/Logs/MemoryLane/` - Windows: `%APPDATA%\MemoryLane\logs\` - Logs contain: capture events, AI calls (without content), errors - No screenshots or semantic content in logs ### Open Source The MemoryLane **desktop client** is open source. - **Repository**: [https://github.com/deusXmachina-dev/memorylane] - Anyone can audit exactly what code runs on your machine - The managed backend service (`api.trymemorylane.com`) used for subscription validation and key provisioning is proprietary ### GDPR / Privacy Regulations - Data minimization: Only semantic data retained long-term; screenshots cleaned up after processing - Local storage: Data stays on user's device - User rights: Access and portability supported via in-app export; deletion requires manual removal of the database file (see Uninstallation section for paths) - Lawful basis: Legitimate interest or consent (via employer) ### APP (Australian Privacy Principles) - APP 5 compliance: Employee notification template provided upon request - Sample 'Employee Information Document' provided upon request ### Employee Notification Before deploying MemoryLane, employers must notify employees. ## Uninstall ### Application Removal 1. **macOS**: Drag MemoryLane to Trash 2. **Windows**: Control Panel → Uninstall a program → MemoryLane ### Data Deletion Standard application removal removes the application binary but **does not automatically delete user data**. To completely remove all data, you must manually delete the following directories: - **macOS**: `~/Library/Application Support/MemoryLane/` and `~/Library/Logs/MemoryLane/` - **Windows**: `%APPDATA%\MemoryLane\` These directories contain: - Local SQLite database (semantic data, OCR text, embeddings) - Configuration files (encrypted API keys, device identity) - Cached embedding model - Application logs ## Contact For security inquiries: **Email**: founders@trymemorylane.com SenseFlow, Inc. 131 Continental Dr, Suite 305 Newark, DE 19713 United States ================================================================================ # Privacy Policy Source: https://trymemorylane.com/privacy ================================================================================ # Privacy Policy **SenseFlow, Inc.** **Effective Date:** January 1, 2026 --- ## Introduction SenseFlow, Inc. ("SenseFlow," "we," "us," or "our"), a Delaware corporation with its principal place of business at 131 Continental Dr, Suite 305, Newark, DE 19713, is committed to protecting your privacy. This Privacy Policy explains how we collect, use, and protect information when you use MemoryLane, our desktop application and related services (collectively, the "Services"). By using our Services, you agree to the practices described in this Privacy Policy. **Core Principle: You only share what you choose to share.** Your data stays on your device unless you explicitly choose to share it, with the following exceptions: (1) the app contacts `api.trymemorylane.com` on startup and periodically to validate your subscription; this request sends your device identifier, and our backend records the request's IP address and derives approximate location (country, city) via a third-party geolocation service for usage analytics and abuse detection (retained up to 90 days); (2) on first use, a local embedding model is downloaded and cached on your device. --- ## Information We Collect ### Screenshots (Ephemeral) MemoryLane captures screenshots of your screen at approximately 1-second intervals during activity sessions (bounded by application switches, 5-second inactivity gaps, and a 5-minute maximum session duration). Screenshots are captured from all visible windows, including incognito/private browser windows, when they are the active window during a capture interval. These screenshots are: - **Processed** to extract semantic meaning (a text description of what was on screen) - **Cleaned up after processing**, a background sweep runs every 10 minutes and removes screenshot and video files older than 1 hour - **Never stored** by SenseFlow or our AI service providers - **Never leave your device** unless you use cloud AI processing (Scenario A), in which case: - **Video mode** (attempted first in the default auto mode): Captured frames are stitched into a 1 FPS MP4 video and transmitted - **Image mode** (fallback): Up to 6 compressed frames per session are transmitted ### Semantic Data (Stored Locally) The AI-extracted descriptions of your screen activity are stored in a local SQLite database on your device. This data: - **Remains on your device** at all times - **Is stored in a local database** on your device, protected by your operating system's access controls. - **Is fully under your control**, you can view and export it anytime. To delete stored activity data, you must manually delete the database file (see Data Retention section for paths). - **Is never transmitted to SenseFlow servers** ### Account Information If you create an account or purchase a subscription, we collect: - Name and email address - Company name (for business accounts) - Payment information (processed by our payment provider) ### Usage Data **The MemoryLane desktop application does not collect any analytics, telemetry, or usage statistics.** The app does not phone home with feature usage, crash reports, or performance metrics. All logging is local only (see local log paths in our [Security & Technical Overview](/security)). **Note:** Our website (trymemorylane.com) does collect analytics via PostHog, see the Cookies and Website section below. **Note:** The app contacts `api.trymemorylane.com` on startup and periodically to validate your subscription. See the Subscription Validation Data section below for details on what is collected. ### Subscription Validation Data When your MemoryLane app validates your subscription, our backend records: your device identifier, the request's IP address, approximate location (country and city, derived via a third-party geolocation service), and the request timestamp. This data is used to (a) validate your subscription status, (b) measure aggregate product usage, and (c) detect licensing abuse. It is retained for up to 90 days and is not linked to your name or email. ### Interaction Monitoring MemoryLane monitors user interaction to detect activity sessions and provide productivity context: | Input Type | What IS Collected | What is NOT Collected | | ------------ | ------------------------------------------------------------------- | ---------------------------------------------------------- | | Keyboard | Session duration, key count, active window title | Keystroke content (what you type) | | Mouse | Click position (x, y coordinates) | Click count (not persisted), what was clicked on (content) | | Scroll | Cumulative scroll amount, direction (vertical/horizontal) | Scroll content | | Window focus | Window title, process name, bundle ID, browser URL, previous window | Window content (captured separately via screenshots) | Interaction data is collected via system-level APIs and native components. Interaction metadata is used ephemerally for session windowing and is not durably stored in the database. It is never transmitted to SenseFlow or any third party. ### Additional Local Data Processing The following processing occurs entirely on your device and never involves data transmission to SenseFlow: - **OCR (Optical Character Recognition)**: On-screen text is extracted locally using macOS Vision framework or Windows native OCR - **Embeddings**: Semantic search vectors are generated locally using an open-source embedding model. On first use, the model is downloaded and cached locally; no user data is sent during this download - **Video stitching**: Screenshot sequences are stitched into short video clips locally via FFmpeg - **Token tracking**: AI API token consumption is tracked locally for usage monitoring ### MCP Server MemoryLane includes an optional Model Context Protocol (MCP) server that allows compatible AI assistants (e.g., Claude, Cursor) to query your local activity data. - The MCP server runs locally via stdio. On first use, the embedding model is downloaded from `huggingface.co` if not already cached; all subsequent communication is local only - You must explicitly configure your AI assistant to connect to the MCP server - SenseFlow has no access to MCP queries or responses - The MCP server can be disabled at any time in app settings --- ## How We Process Your Data MemoryLane offers three AI processing options. What happens to your data depends on which option you or your organization chooses: ### Scenario A: OpenRouter (Default) When using the default cloud processing, your data goes **directly from your device to OpenRouter**, SenseFlow does not proxy or route your data: 1. Screenshots are captured on your device during activity sessions 2. Depending on the capture mode: - **Auto mode** (default): Attempts video first and falls back to image mode if video processing is unavailable or fails. - **Video mode**: Captured frames are stitched into a 1 FPS **video** via FFmpeg. The video is sent alongside activity metadata for summarization. - **Image mode**: Up to 6 key screenshots are sampled and sent as images for summarization. 3. The video or images are sent via **TLS-encrypted connection** directly to OpenRouter API. OpenRouter load-balances and geo-balances requests across AI providers. 4. OpenRouter routes to the appropriate AI model chain: - **Video mode**: AI models via ZDR-compliant provider endpoints - **Image mode**: AI models via ZDR-compliant provider endpoints - **Pattern detection (Pro)**: AI model via ZDR-compliant endpoint, receives activity metadata including text summaries, window titles, app names, domain names, and timestamps (no images) 5. **Zero Data Retention (ZDR)** is enabled at the OpenRouter account level for managed API keys. When ZDR is enabled, OpenRouter only routes to ZDR-compliant endpoints. If you bring your own OpenRouter key, you are responsible for enabling ZDR in your account settings. 6. Only semantic data (text) returns to your device 7. Screenshots and video files are **cleaned up after processing**; a background sweep removes files older than 1 hour **Sub-processors for Scenario A:** - OpenRouter (AI API routing and processing), ZDR enabled for managed keys. OpenRouter routes requests to ZDR-compliant AI provider endpoints (e.g., Google Vertex for Gemini models). See [OpenRouter's ZDR provider list](https://openrouter.ai/docs/guides/features/zdr) for current downstream providers. - Stripe (payment processing, via browser redirect, no screen data) - Hugging Face (one-time embedding model download, no user data transmitted) - GitHub (application update checks and downloads) - ipinfo.io (IP geolocation for subscription validation, receives the IP address of your validation request) ### Scenario B: Your Cloud AI Endpoint When you configure your own AI endpoint: - Screenshots are sent to **your chosen endpoint** - SenseFlow **never sees your data** - You are responsible for your endpoint's privacy practices ### Scenario C: Fully Local Models When using local AI models: - **No data ever leaves your device** - Processing happens entirely on your machine - Maximum privacy and air-gap capability --- ## What We Do NOT Collect SenseFlow does not: - Store your screenshots on our servers - Access the content of your semantic data - Record keystroke **content** (what you type), MemoryLane monitors typing session duration and key count for activity detection only, via uiohook-napi - Record audio - Collect any analytics, telemetry, or usage statistics **in the desktop application**, all logging is local only (our website uses PostHog for analytics; see Cookies and Website section) - Capture incognito/private browser windows by default, MemoryLane pauses capture while a private/incognito browser window is the **active (focused)** window, and applies the same rule to any apps or websites on your exclusion list. **Caveat:** capture is gated by the active window only. On a large display with several apps visible side-by-side, if the active window is _not_ excluded, the screenshot captures the full screen, so a private browser or other sensitive app next to the active window may still appear in the image. Close or minimize sensitive windows before switching focus to a non-excluded app --- ## Data Sharing ### We Do Not Sell Your Data We do not sell, rent, or trade your personal information to third parties. ### Service Providers We work with service providers to operate our business: | Provider | Purpose | Data Accessed | | ------------ | ------------------------------------------ | ---------------------------------------------------------------------------------------------------------------------------------------------------------------- | | OpenRouter | AI API routing and processing (Scenario A) | Video or screenshots, activity metadata. Routes to ZDR-compliant provider endpoints (see [OpenRouter ZDR list](https://openrouter.ai/docs/guides/features/zdr)). | | Stripe | Payment processing | Payment and subscription info (via browser redirect) | | Hugging Face | Embedding model hosting | Model download request only, no user data | | GitHub | Application updates | Update check requests | ### Legal Requirements We may disclose information if required by law, regulation, or legal process. ### Business Transfers If SenseFlow is involved in a merger, acquisition, or sale, your information may be transferred as part of that transaction. We will notify you of any such change. --- ## Workplace Use and Employee Privacy MemoryLane is designed for business use. When your employer deploys MemoryLane: - Your employer is the **data controller** for any personal data processed - SenseFlow acts as a **data processor** or service provider - Your employer must provide you with notice before deployment (see our Employee Notice template) - You retain control over your local data, you can pause, view, and export your data **Employees**: If you have questions about how your employer uses MemoryLane, please contact your employer's IT or HR department. --- ## Your Rights and Choices ### Access Your Data All your semantic data is stored locally on your device. You can view it anytime within the MemoryLane app. ### Delete Your Data Screenshot and video files are automatically cleaned up after processing (background sweep removes files older than 1 hour). To delete stored activity data (semantic summaries, OCR text, and embeddings), you must manually delete the SQLite database file: - **macOS**: `~/Library/Application Support/MemoryLane/memorylane.db` - **Windows**: `%APPDATA%\MemoryLane\memorylane.db` You can also delete your API key and reset capture settings through the app's interface. ### Export Your Data You can export your local database as a ZIP file containing the SQLite database. This requires SQLite-compatible tooling to read. ### Pause Collection You can pause screenshot capture at any time with one click. No data is collected while paused. ### Uninstall You can uninstall MemoryLane at any time. Standard OS uninstallation removes the application binary but **does not automatically delete user data**. To completely remove all data, you must manually delete the application data directories: - **macOS**: `~/Library/Application Support/MemoryLane/` and `~/Library/Logs/MemoryLane/` - **Windows**: `%APPDATA%\MemoryLane\` ### No Desktop App Analytics The MemoryLane desktop application does not collect any analytics, telemetry, or usage statistics. No data is sent to SenseFlow for analytics purposes from the app. Our website (trymemorylane.com) does use PostHog for analytics, see the Cookies and Website section below. --- ## Data Security We implement appropriate security measures to protect your information: - **Credential encryption**: API keys and device ID are encrypted via Electron safeStorage (uses macOS Keychain / Windows DPAPI to derive encryption keys; encrypted data stored in local config files) - **Local database**: SQLite, unencrypted; protected by OS-level file permissions and user account access controls - **Encryption in transit**: All API communications use HTTPS/TLS (TLS version negotiated by the OS and server) - **Zero Data Retention**: ZDR is enabled at the OpenRouter account level for managed API keys (see our [Security & Technical Overview](/security)) - **Open source**: Our desktop client code is publicly auditable. The managed backend service (`api.trymemorylane.com`) is proprietary --- ## Data Retention - **Screenshots and video files**: Cleaned up after processing; background sweep removes files older than 1 hour - **Semantic data, OCR text, and embeddings**: Stored locally in the SQLite database indefinitely. No in-app deletion interface currently exists; to delete, manually remove the database file (see paths in Delete Your Data section above) - **Interaction metadata**: Used ephemerally for session windowing; not durably stored in the database - **Account information**: Retained while your account is active and for a reasonable period thereafter for legal and business purposes --- ## International Users SenseFlow is based in the United States. If you use our Services from outside the US: - Your data processing (Scenario A) occurs in the US and/or regions where our AI providers operate - Local storage on your device remains in your jurisdiction - We comply with applicable data protection laws, including GDPR for EU users ### For EU/EEA Users Under GDPR, you have additional rights including: - Right to access, rectification, and erasure - Right to data portability - Right to object to processing - Right to lodge a complaint with a supervisory authority To exercise these rights, contact us at founders@trymemorylane.com. --- ## Children's Privacy MemoryLane is not intended for use by children under 18. We do not knowingly collect information from children. If you believe a child has provided us with personal information, please contact us. --- ## Cookies and Website Our website (trymemorylane.com) uses cookies and analytics: - **Essential functionality**: Required for site operation - **Analytics**: We use PostHog to collect website usage data including page views, visitor interactions, and referral sources. PostHog may set cookies to distinguish unique visitors and track sessions. This data helps us understand how visitors use our website and improve our marketing. - **Preference storage**: Storing your site preferences You can control cookies through your browser settings. Website analytics data is separate from the MemoryLane desktop application, the desktop app does not collect or transmit any analytics. --- ## Changes to This Policy We may update this Privacy Policy from time to time. We will notify you of material changes by: - Posting the updated policy with a new effective date - Sending an email to account holders (for significant changes) Your continued use of the Services after changes constitutes acceptance of the updated policy. --- ## Contact Us If you have questions about this Privacy Policy or our privacy practices: **Email:** founders@trymemorylane.com **Mail:** SenseFlow, Inc. 131 Continental Dr, Suite 305 Newark, DE 19713 United States **Website:** trymemorylane.com ================================================================================ # Terms and Conditions Source: https://trymemorylane.com/terms ================================================================================ # Terms and Conditions **SenseFlow, Inc.** **Effective Date:** January 1, 2026 --- ## 1. Agreement Overview These Terms and Conditions ("Terms") form a legally binding agreement between you ("Customer" or "you") and SenseFlow, Inc. ("SenseFlow," "we," or "us"), a Delaware corporation with its principal place of business at 131 Continental Dr, Suite 305, Newark, DE 19713, governing your access to and use of MemoryLane and related services. By accessing our platform, creating an account, or using our services, you agree to these Terms. If you disagree, you must not use our services. These Terms include a binding arbitration provision that affects your legal rights. --- ## 2. Definitions - **Account**: Your registered account that enables access to our Services. - **Authorized Users**: Individuals you authorize to access the Services through your Account. - **Customer Data**: Screenshots captured, video clips stitched, semantic data extracted, OCR text, embeddings, interaction metadata (keyboard session durations, key counts, mouse click positions, scrolls, used ephemerally, not stored in database), window context (titles, process names, bundle identifiers, URLs), pattern analysis data, and any other content processed through MemoryLane on your devices. - **Platform**: The MemoryLane desktop application, website, and related technologies. - **Services**: All software, features, and functionality provided through the Platform, including AI processing options. - **Subscription**: The service plan you purchase that determines your level of access. - **Subscription Fees**: The fees you pay for your Subscription. --- ## 3. Services MemoryLane is a desktop application that captures screenshots, processes them using AI to extract semantic meaning, monitors user interaction metadata (keyboard sessions, mouse clicks, scrolls, no keystroke content), collects window context (titles, process names, URLs), performs local OCR and embedding generation, and stores the resulting data locally on your device. MemoryLane also includes an optional MCP (Model Context Protocol) server for local AI assistant integration and pattern detection for productivity insights (Pro feature). We offer three AI processing scenarios: - **Scenario A (OpenRouter)**: AI processing via OpenRouter directly from your device, SenseFlow does not proxy your data. OpenRouter routes requests to ZDR-compliant AI provider endpoints (e.g., Google Vertex for Gemini models). The default auto mode attempts video processing first (captured frames stitched into a 1 FPS MP4) and falls back to image mode (up to 6 sampled screenshots). SenseFlow configures Zero Data Retention (ZDR) at the OpenRouter account level for managed API keys; BYOK users are responsible for their own settings. See [OpenRouter's ZDR provider list](https://openrouter.ai/docs/guides/features/zdr) for current downstream providers. - **Scenario B (Your Cloud Endpoint)**: You provide and manage your own AI endpoint - **Scenario C (Local Models)**: AI processing runs entirely on your device We may update the Services periodically. These Terms apply to all updates. --- ## 4. Your Account ### 4.1 Account Eligibility To use our Services, you must: - Be at least 18 years old - Be legally competent to enter into binding contracts - Provide accurate information when creating an Account - Use the Services for business purposes ### 4.2 Account Security You are responsible for: - Maintaining the confidentiality of your login credentials - All activities that occur under your Account - Immediately notifying us at founders@trymemorylane.com of any unauthorized use or security breach ### 4.3 Authorized Users If you authorize others to use the Services under your Account, you are responsible for: - Ensuring Authorized Users comply with these Terms - Ensuring Authorized Users do not share access with unauthorized parties - Disabling access for Authorized Users who should no longer have access - All actions taken by your Authorized Users --- ## 5. Restrictions and Acceptable Use You agree not to: - Reverse engineer, decompile, or disassemble any part of the Services - Sell, resell, license, sublicense, distribute, rent, or lease the Services - Use the Services for any illegal purpose - Attempt to gain unauthorized access to the Services or related systems - Interfere with the proper functioning of the Services - Upload harmful content or introduce viruses or malicious code - Use the Services to infringe on others' intellectual property rights - Use the Services in violation of applicable export control or sanctions laws - Circumvent any security measures or access controls - Use the Services to monitor individuals without proper notice and consent --- ## 6. License Grant SenseFlow grants you a limited, non-exclusive, non-transferable, revocable license to: - Download and install MemoryLane on devices you own or control - Use the Services according to your Subscription plan - Access our documentation and support materials This license does not grant any rights to SenseFlow's intellectual property except as explicitly stated in these Terms. --- ## 7. Intellectual Property ### 7.1 SenseFlow's Rights SenseFlow owns and retains all rights, title, and interest in the Services, including all intellectual property rights. Nothing in these Terms transfers ownership rights to you. ### 7.2 Customer Data Rights **You own your data.** - Screenshots, video clips, semantic data, OCR text, embeddings, interaction metadata, window context, pattern data, and activity summaries processed by MemoryLane remain your property - Data is stored locally on your device, under your control - We do not claim any ownership rights to your Customer Data You grant SenseFlow a limited license to process your Customer Data solely to provide the Services (e.g., sending compressed screenshots and video clips to AI providers for processing under Scenario A, processing payments via Stripe, and generating activity summaries for pattern detection). ### 7.3 Open Source Components MemoryLane includes open source software components. The applicable open source licenses govern your use of those components. ### 7.4 Feedback If you provide feedback or suggestions about our Services, you grant us a perpetual, royalty-free license to use and incorporate such feedback without compensation. --- ## 8. Data Protection and Privacy ### 8.1 Privacy Policy Our [Privacy Policy](/privacy) is incorporated into these Terms and describes how we handle your information. ### 8.2 Data Processing - **Scenario A**: Your device sends data directly to OpenRouter for AI processing; SenseFlow acts as data processor by provisioning API keys and selecting providers; you remain the data controller. The app also contacts `api.trymemorylane.com` on startup with a device identifier for subscription validation. - **Scenario B/C**: SenseFlow provides software only; you control all data processing For enterprise deployments requiring a Data Processing Agreement, see our DPA. ### 8.3 Employee Notification If you deploy MemoryLane in a workplace setting, you are responsible for providing appropriate notice to employees before capture begins. See our Employee Notice template. ### 8.4 Security Measures We implement appropriate technical and organizational security measures. See our [Security & Technical Overview](/security) for details. --- ## 9. Payment Terms ### 9.1 Fees and Payment You agree to pay all applicable Subscription Fees according to your chosen plan. Fees are due as specified in your invoice or at time of purchase. ### 9.2 Taxes All fees are exclusive of taxes. You are responsible for all applicable taxes except for taxes based on SenseFlow's income. ### 9.3 Automatic Renewal Subscriptions automatically renew for the same period unless canceled before the renewal date. You authorize us to charge the payment method on file for renewals. ### 9.4 No Refunds Subscription fees are non-refundable. We do not provide prorated refunds for canceled Subscriptions. ### 9.5 Late Payments Overdue payments may accrue interest at 1.5% per month or the maximum rate permitted by law. We may suspend access to Services for accounts more than 15 days past due. --- ## 10. Term and Termination ### 10.1 Term These Terms remain in effect throughout your use of the Services. ### 10.2 Termination by You You may terminate your Subscription at any time by: - Canceling through your account settings - Contacting us at founders@trymemorylane.com - Uninstalling the application ### 10.3 Termination by SenseFlow We may terminate or suspend your access to the Services: - If you breach these Terms - If you fail to pay Subscription Fees when due - If we are required to do so by law - If we cease offering the Services (with reasonable notice) ### 10.4 Effect of Termination Upon termination: - Your right to access the Services ends - You should uninstall the application - Your locally stored data remains on your device (you control its deletion by manually removing the application data directory) - We will delete your account information within 30 days ### 10.5 Survival Provisions related to intellectual property rights, warranties, liability, indemnification, and dispute resolution survive termination. --- ## 11. Warranties and Disclaimers ### 11.1 Limited Warranty SenseFlow warrants that the Services will perform substantially in accordance with our documentation under normal use. ### 11.2 Disclaimer EXCEPT AS EXPRESSLY PROVIDED HEREIN, THE SERVICES ARE PROVIDED "AS IS" AND "AS AVAILABLE" WITHOUT WARRANTIES OF ANY KIND, WHETHER EXPRESS, IMPLIED, STATUTORY, OR OTHERWISE. SENSEFLOW SPECIFICALLY DISCLAIMS ALL IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, TITLE, AND NON-INFRINGEMENT. SENSEFLOW DOES NOT WARRANT THAT THE SERVICES WILL BE UNINTERRUPTED, ERROR-FREE, OR MEET YOUR SPECIFIC REQUIREMENTS. --- ## 12. Limitation of Liability ### 12.1 Exclusion of Damages TO THE MAXIMUM EXTENT PERMITTED BY LAW, SENSEFLOW WILL NOT BE LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, OR PUNITIVE DAMAGES, INCLUDING LOST PROFITS, LOST DATA, OR BUSINESS INTERRUPTION, ARISING OUT OF OR RELATED TO THESE TERMS OR THE SERVICES. ### 12.2 Cap on Liability IN NO EVENT WILL SENSEFLOW'S TOTAL LIABILITY ARISING OUT OF OR RELATED TO THESE TERMS OR THE SERVICES EXCEED THE TOTAL AMOUNT PAID BY YOU FOR THE SERVICES IN THE TWELVE (12) MONTHS PRECEDING THE EVENT GIVING RISE TO THE CLAIM. ### 12.3 Exceptions These limitations do not apply to: - SenseFlow's willful misconduct or gross negligence - Death or personal injury caused by SenseFlow's negligence - Any other liability that cannot be excluded under applicable law --- ## 13. Indemnification You agree to indemnify and hold harmless SenseFlow and its affiliates, officers, directors, employees, and agents from any claims, liabilities, damages, losses, costs, or expenses (including reasonable attorneys' fees) arising from: - Your use of the Services - Your Customer Data - Your violation of these Terms - Your violation of any third-party rights - Your failure to provide required employee notifications --- ## 14. Dispute Resolution ### 14.1 Informal Resolution Before initiating formal proceedings, you agree to contact us at founders@trymemorylane.com to attempt informal resolution. We will try to resolve disputes within 30 days. ### 14.2 Arbitration Any dispute arising from these Terms that cannot be resolved informally will be resolved by binding arbitration administered by the American Arbitration Association ("AAA") under its Commercial Arbitration Rules. The arbitration will be conducted in English, and judgment on the award may be entered in any court of competent jurisdiction. ### 14.3 Location Arbitration will take place in Wilmington, Delaware, or another mutually agreed location. Either party may participate remotely. ### 14.4 Exceptions Either party may seek injunctive relief in a court of competent jurisdiction or file claims in small claims court. ### 14.5 Class Action Waiver YOU AGREE TO RESOLVE DISPUTES WITH SENSEFLOW ON AN INDIVIDUAL BASIS AND WAIVE ANY RIGHT TO PARTICIPATE IN A CLASS ACTION, CLASS ARBITRATION, OR REPRESENTATIVE PROCEEDING. --- ## 15. Governing Law These Terms are governed by the laws of the State of Delaware, United States, without regard to conflict of law principles. The United Nations Convention on Contracts for the International Sale of Goods does not apply. --- ## 16. Miscellaneous ### 16.1 Entire Agreement These Terms, together with our Privacy Policy and any applicable DPA, constitute the entire agreement between you and SenseFlow regarding the Services. ### 16.2 Severability If any provision of these Terms is found unenforceable, that provision will be modified to the minimum extent necessary, and the remaining provisions will remain in effect. ### 16.3 No Waiver Failure to enforce any provision of these Terms does not constitute a waiver of that or any other provision. ### 16.4 Assignment You may not assign these Terms without SenseFlow's prior written consent. SenseFlow may assign these Terms to an affiliate or successor entity. ### 16.5 Force Majeure Neither party is liable for delays or failures due to causes beyond its reasonable control, including natural disasters, pandemic, terrorism, war, or service failures. ### 16.6 Notices Notices to SenseFlow must be sent to founders@trymemorylane.com or our physical address. Notices to you may be sent to the email address associated with your Account. ### 16.7 Export Compliance The Services may be subject to export laws and regulations. You agree to comply with all applicable export laws. --- ## 17. Contact Information If you have questions about these Terms: **Email:** founders@trymemorylane.com **Mail:** SenseFlow, Inc. 131 Continental Dr, Suite 305 Newark, DE 19713 United States **Website:** trymemorylane.com ================================================================================ # How much does process mapping cost Source: https://trymemorylane.com/blog/how-much-does-process-mapping-cost ================================================================================ # How much does process mapping cost **Contents** - [Three pricing models](#three-pricing-models) - [Free and low-cost process mapping tools](#free-and-low-cost-process-mapping-tools) - [Consultant pricing](#consultant-pricing) - [Automated process mapping software](#automated-process-mapping-software) - [The hidden costs](#the-hidden-costs) - [Which option fits your budget](#which-option-fits-your-budget) - [FAQ](#faq) Process mapping cost depends on how you do it. A whiteboard and a marker cost nothing. A Big Four consulting engagement costs six figures. Automated process mapping software sits somewhere in between, and for most teams, delivers better results than either extreme. In this post I break down what each approach costs, what you get for the money, and where the hidden expenses might be. ## Three pricing models There are three ways to map your processes with three different price tags: 1. **Do it yourself with cheap tools** (starts at $20/user/month) 2. **Hire consultants** (starts at $200,000) 3. **Use automated process mapping software** (starts at $100/user/month) The right choice depends on your team size, how many processes you need to map, and whether you need a one-time snapshot or continuous visibility. ## Free and low-cost process mapping tools Most teams start here. You pick a diagramming or whiteboard tool, interview your team, and draw flowcharts. The software cost is low. The time cost is high. ### What is available **Free options** - **[Miro](https://miro.com/)** has a free tier with 3 boards. Good for brainstorming and rough workflow mapping with your team on a shared canvas. - **[Lucidchart](https://www.lucidchart.com/)** offers a free tier with limited documents. Better structured than Miro for formal diagrams. - **[Google Drawings](https://docs.google.com/drawings)** is free with any Google account. Basic, but it works for simple flowcharts. - **[draw.io (diagrams.net)](https://www.diagrams.net/)** is fully free and open source. Open source process mapping at its best. No account required, runs in the browser, exports to most formats. **Paid options** - **[Miro](https://miro.com/) Starter** at $8/user/month. Removes board limits and adds collaboration features. - **[Lucidchart](https://www.lucidchart.com/)** paid plan at $7.95/month. Full diagramming with BPMN support. - **[Microsoft Visio](https://www.microsoft.com/en-us/microsoft-365/visio/flowchart-software)** at $5 to $15/user/month (requires Microsoft 365). The enterprise standard for formal BPMN process mapping. For a full breakdown of each tool, see our [process mapping software comparison](https://trymemorylane.com/blog/best-process-mapping-software). ### What you actually spend The software is cheap, but your team's time needs to be priced in. Someone on your team has to run the initiative and schedule interviews, synthesize what people describe, draw diagrams, get feedback, iterate and all this for a likely imperfect result. For a team of 20 with 10 core workflows, budget two or more full weeks of project time. At a fully loaded cost of $75/hour, that is $3,000 to $6,000 in internal labor on top of whatever the software and opportunity cost is. And you still end up with an incomplete picture, because people cannot accurately describe work they do on autopilot. ## Consultant pricing Consultants bring structure and experience. They know how to run workshops, they have frameworks, and they deliver polished reports. The output is better than most DIY efforts. You will no doubt be in good hands. But that comes with a hefty cost. ### What it costs - **Boutique process consultants** run $100,000 or more for a scoped engagement (4 to 6 weeks, focused on only one department or function). - **Big Four firms** (Deloitte, EY, PwC, KPMG) start at $250,000+ under their "digital transformation" or "intelligent automation" practice. [Gartner estimates](https://www.gartner.com/en/documents/4012770) the average process mining initiative at $250K+ at the enterprise level. - **Freelance process analysts** charge $25,000 or more for a narrowly targeted project you make a set of assumptions from. Harder to find, but good value if you find someone competent. ### What you get A polished deliverable with process maps, gap analysis, recommendations. The method is solid. But the output is still interview-based and extrapolated from a few team members to entire departments. That means it's full of assumptions, not ground-truth. Consultants ask your team how they work, observe them for a few days, and synthesize. They miss the same things your team misses – tab switches, copy-paste sequences, multi-step workarounds nobody mentions because they barely register. The other problem is shelf life. A consultant maps your processes in Q3, and by Q4 the map is already wrong. Teams change, tools change, workarounds evolve. You paid for a snapshot, not ongoing visibility. ## Automated process mapping software This is the newest category and most interesting category. Software observes how work actually happens on people's screens and maps the workflows without anyone having to describe anything. Instead of asking what people do, you see what they actually did. ### Enterprise tools The pioneers here are [UiPath](https://www.uipath.com/product/process-mining), [Celonis](https://www.celonis.com/), and [Mimica](https://www.mimica.ai/). They built excellent products for large organizations with thousands of employees. The cost matches that. $100,000+ per year, multi-month implementations, dedicated IT support. For companies with 5,000 employees processing millions of transactions, the ROI is there. For a company with 50 or 100 people, not really. We compared these tools in detail in [MemoryLane vs UiPath, Celonis, and Mimica](https://trymemorylane.com/blog/memorylane-vs-uipath-celonis-mimica). ### MemoryLane automates process mapping for mid-segment We built [MemoryLane](https://trymemorylane.com) to bring automated process mapping to small and mid-sized companies. It installs in [five minutes](https://trymemorylane.com/guide), runs in the background for a few weeks, and produces a ranked list of workflows sorted by time spent and automation potential. No interviews. No flowcharts drawn from memory. Your data stays with you. The [self-serve plan](https://trymemorylane.com/#pricing) starts at $100/month for a single user, so it's easy to trial. Team deployments are priced per organization. If you are wondering about the privacy side, read [why task mining is not employee monitoring](https://trymemorylane.com/blog/task-mining-vs-employee-monitoring). ## The hidden costs The sticker price of process mapping is misleading. The real money goes to three places most teams do not budget for. ### Time The biggest cost in any process mapping project is really the time your team spends away from their actual work. Interviews, workshops, reviews, feedback loops. For a consultant engagement, your team's involvement alone can cost as much as the consultant's fee. Workflow mapping software that runs automatically eliminates 90% of this. It works in the background and nobody needs to change how they work. ### Accuracy An inaccurate process map is worse than no map at all. It gives you false confidence. You automate the wrong thing, build the wrong SOP, or invest in the wrong area. Manual methods capture the main steps but miss details like copy-pasting between systems. Screen-level observation captures everything. ### Maintenance A process map is only useful if it reflects how work happens today. Manual maps go stale immediately. Consultant deliverables expire within a quarter. Software that keeps observing keeps the data current. If you factor in the cost of re-mapping every 6 months with a consultant ($100K+ each time) versus continuous monitoring with software (starting at $100/user/month), the software pays for itself within the first cycle. ## Which option fits your budget | Approach | Upfront cost | Ongoing cost | Best for | | ---------------------- | --------------- | ---------------------- | ------------------------------------------------------------------ | | Free tools + DIY | $20+/user/month | Your team's time | Teams under 10 with simple workflows | | Freelance analyst | $25,000+ | $25K+ per refresh | One-time documentation of a specific department | | Boutique consultant | $100,000+ | $100K+ per refresh | Mid-size companies needing structured analysis | | Big Four engagement | $250,000+ | Similar per engagement | Large enterprises with complex operations | | Enterprise task mining | $100,000+/year | Same | Enterprises with thousands of employees | | MemoryLane | $100/user/month | Same | Teams that want continuous visibility without the enterprise price | For most small and mid-size teams, the practical choice is between DIY (free tools + your time) and automated software ($100/user/month + minimal time). The consultant path makes sense when you have the budget and need a human to interpret the findings and present to the board. ## FAQ ### Is there free process mapping software that actually works? Yes. [draw.io](https://www.diagrams.net/) is fully free, open source, and handles flowcharts and BPMN diagrams well. [Miro](https://miro.com/) and [Lucidchart](https://www.lucidchart.com/) have usable free tiers. The limitation is that all free tools only draw diagrams. They do not help you discover what to draw. You still need to figure out what your processes are through interviews and observation. ### How long does a process mapping project take? DIY with free tools takes 2 to 4 weeks for 5 to 10 workflows, if someone owns it full-time. Consultants need 4 to 12 weeks depending on scope. Automated process mapping tools like MemoryLane produce better results in the same time window, but normally generate 30+ detailed worklows grounded in objective data. ### Can I map processes without interviews? Yes. Automated process mapping software like MemoryLane observes screen-level activity and identifies workflows without anyone describing them. This is called [task mining](https://trymemorylane.com/blog/task-mining-for-small-teams). It captures work that people do on autopilot and would never think to mention in an interview. ### What does end-to-end process mapping mean? End-to-end mapping means documenting a workflow from its trigger to its final output, across every system and step involved. Most manual mapping captures the main steps but misses the in-between work. Tab switches, email follow-ups, data validation, copy-paste between systems. Automated tools capture this full chain because they observe everything that happens on screen, whether someone remembers it or not. ## Next steps Try [MemoryLane](https://trymemorylane.com) on the [self-serve plan](https://trymemorylane.com/#pricing) to see what it finds on your own machine first. Read more about [how to map your processes](https://trymemorylane.com/blog/how-to-map-your-processes) if you are still weighing your options. [Get in touch](mailto:founders@trymemorylane.com) if you need help with team deployment or a customized analysis. ================================================================================ # Best process mapping software in 2026 Source: https://trymemorylane.com/blog/best-process-mapping-software ================================================================================ # Best process mapping software in 2026 **Contents** - [Why process maps matter](#why-process-maps-matter) - [How we evaluated](#how-we-evaluated) - [Microsoft Visio](#microsoft-visio) - [Miro](#miro) - [Lucidchart](#lucidchart) - [Scribe](#scribe) - [MemoryLane](#memorylane) - [Comparison](#comparison) - [Which one should you use](#which-one-should-you-use) - [FAQ](#faq) Most process mapping projects die the same way. Someone opens a diagramming tool, stares at a blank canvas, realizes they need to interview twelve people before drawing a single box, and two weeks later the flowcharts are half done and the person who started them got pulled into something else. We have watched this happen at companies of every size. Whether you are shopping for business process mapping software or just trying to document a single workflow, the tool you pick determines whether you spend your time drawing boxes or improving the processes those boxes represent. ## Why process maps matter Most companies do not have written SOPs (Standard Operating Procedures). We ask every new customer at kickoff. The answer is almost always no. The processes live in people's heads, and what people describe in interviews rarely matches what they actually do. That gap is where the waste hides. The invoice copied between three systems. The approval that takes four emails instead of one click. The weekly report someone builds by hand from five browser tabs. The sales process mapping that never happened because the pipeline "works fine," until someone checks the KPIs and finds half the steps are manual. You cannot fix what you cannot see, and a process map is how you start seeing. Our post on [how to map your processes](https://trymemorylane.com/blog/how-to-map-your-processes) covers the three approaches in detail: doing it yourself, hiring consultants, and automated discovery. ## How we evaluated We rated each business mapping software option on four dimensions: 1. **Manual vs. automated discovery.** Does the tool help you find processes, or do you need to know what to map before you start? 2. **Learning curve.** How quickly can a non-technical team member produce useful output? 3. **Export and interoperability.** Can you get your maps out as BPMN, Mermaid, or other formats that feed into automation tools? 4. **Pricing for small and mid-size teams.** Enterprise tools with six-figure minimums exist, but they are not options for a 50-person company. We focused on what teams under 500 people can actually buy. For a full cost breakdown across all approaches, see [how much does process mapping cost](https://trymemorylane.com/blog/how-much-does-process-mapping-cost). ## The tools ### Microsoft Visio [Visio](https://www.microsoft.com/en-us/microsoft-365/visio/flowchart-software) has been around since the 1990s and remains the default in large organizations, mostly because it ships with Microsoft 365 enterprise licenses. Visio does formal flowcharts well. BPMN process mapping is its core strength: notation-standard shapes, a large library, deep integration with SharePoint and Teams. If your company already runs Microsoft and needs BPMN diagrams, Visio checks the box. The learning curve is steep. The interface feels dated. And the core limitation applies to every drawing tool on this list: Visio does not help you discover what to draw. You still need to do all the process discovery yourself through interviews and guesswork. **Pricing:** Plan 1 at $5/user/month (browser). Plan 2 at $15/user/month (desktop app). Both require Microsoft 365. **Best for:** Large enterprises on Microsoft that need formal BPMN or UML diagrams. ### Miro [Miro](https://miro.com/) is a collaborative whiteboard, not a dedicated process mapping tool. Miro process mapping is popular because the tool is easy to pick up and everyone can edit at the same time. For brainstorming and early-stage discovery, Miro works well. Run a virtual workshop, drop sticky notes, draw rough flows, iterate in real time. The problem is that Miro is a canvas. Your flowcharts are drawings, not structured data. No BPMN export. No way to link a process to its owner or review date. Once you have more than a dozen processes mapped, the boards get messy. And you are still drawing from what people tell you, which means you are still missing the work nobody thinks to mention. **Pricing:** Free (3 boards). Starter at $8/user/month. Business at $16/user/month. Enterprise on request. **Best for:** Collaborative brainstorming and rough process sketches. ### Lucidchart [Lucidchart](https://www.lucidchart.com/) sits between Visio and Miro. Proper diagramming with BPMN support, but browser-based and modern. Where Lucid pulls ahead is the process management layer they have been building on top. Link diagrams to data, create a central process repository, manage ownership and versioning. For teams that want one place to document and maintain all their processes, this is the strongest option among the drawing tools. The learning curve lands in the middle. Easier than Visio, more complex than Miro, because Lucid is trying to be both a drawing tool and a repository. The paid plans add up once you need collaboration and admin features. And the mapping itself still depends on interviews. You are building the map by hand. **Pricing:** Free (limited). Single user at $7.95/month. Team at $9/user/month. Business and Enterprise on request (adds process repository, SSO, admin controls). **Best for:** Structured process documentation with a central repository, built manually. ### Scribe [Scribe](https://scribehow.com/) flips the approach. Instead of drawing a flowchart, you record your screen while doing the work. Scribe generates a step-by-step guide with screenshots and annotations automatically. Click record, do the task, click stop. The output is a visual SOP anyone can follow, and the learning curve is the lowest on this list. For creating training materials and documentation, Scribe is hard to beat. The limitation is scope. Scribe captures one process at a time, and you pick which one. It does not observe your work passively or tell you which processes take the most time. You need to know what to map before you can map it. Automated documentation, not automated discovery. **Pricing:** Free (unlimited users, limited guides). Pro at $23/user/month (unlimited guides, branding, analytics). Enterprise on request. **Best for:** Clean SOPs and training docs when you already know which processes to document. ### MemoryLane We built [MemoryLane](https://trymemorylane.com) because every tool above leaves the hardest part to you: figuring out what to map. MemoryLane runs in the background and observes how work actually happens. Apps, windows, repeated steps, over a normal work week. The output is a ranked list of workflows sorted by time spent and automation potential, with KPI data you can act on: hours per week, frequency, estimated cost. No interviews. No one decides what to record. The tool finds the processes. From there, export as BPMN, Mermaid, PDF, or PNG (all coming soon) and feed the results into whichever automation tool you use. This is [task mining](https://trymemorylane.com/blog/task-mining-for-small-teams) in practice. The enterprise players (UiPath, Celonis, Mimica) do similar things, but start at $100K+ and require dedicated implementation teams. See our [comparison](https://trymemorylane.com/blog/memorylane-vs-uipath-celonis-mimica) for the full breakdown. Data stays on the device. Screenshots are processed and immediately deleted through [zero-data-retention](https://trymemorylane.com/security) AI providers. On-premise deployment is available for tighter controls. The privacy model is covered in [why task mining is not employee monitoring](https://trymemorylane.com/blog/task-mining-vs-employee-monitoring). **Pricing:** Self-serve starts at [$100/month](https://trymemorylane.com/#pricing). Team plans available. **Best for:** Finding out where time goes before deciding what to automate. ## Comparison | | Visio | Miro | Lucidchart | Scribe | MemoryLane | | ---------------------- | ---------------------- | ------------- | -------------------------------- | ------------------- | --------------------------------------------- | | **Type** | Diagramming | Whiteboard | Diagramming + repo | Screen recording | Automated discovery | | **Discovery** | Manual | Manual | Manual | Semi-automated | Fully automated | | **Learning curve** | High | Low | Medium | Very low | Very low | | **BPMN support** | Yes | No | Yes | No | Yes (export) | | **Process repository** | No | No | Yes (paid) | No | Built-in | | **Shows time spent** | No | No | No | No | Yes | | **Starting price** | $5/user/mo | Free | Free | Free | [$100/mo](https://trymemorylane.com/#pricing) | | **Best for** | Enterprise diagramming | Brainstorming | Enterprise process documentation | Manual SOP creation | Automated process discovery | ## Which one should you use | If you want to... | Use | | ------------------------------------------------------- | ------------------------------------------- | | Draw formal BPMN diagrams on Microsoft | **Visio** | | Brainstorm workflows on a shared whiteboard | **Miro** | | Build a structured process documentation repository | **Lucidchart** | | Create step-by-step SOPs for processes you already know | **Scribe** | | Find out where time actually goes and what to automate | **[MemoryLane](https://trymemorylane.com)** | Most teams we work with combine tools. MemoryLane finds and ranks the processes. Then they export the top ones into Lucidchart or Scribe for documentation, or feed them directly into their automation stack. ## FAQ ### What is the difference between process mapping and process mining? Process mapping means documenting how work happens, usually as flowcharts or SOPs. Process mining analyzes event logs from enterprise systems (SAP, Oracle) to reconstruct transaction flows. It requires structured system data and typically costs $200K+. Task mining, which MemoryLane uses, observes screen-level activity instead and works without any system integration. ### Is Microsoft Visio still worth it in 2026? If your organization already pays for Microsoft 365 and needs formal BPMN or UML diagrams, Visio still works. Starting from scratch, browser-based tools like Lucidchart offer the same diagramming with a shorter learning curve and real-time collaboration. Visio's main advantage is the bundled license. ### What is BPMN? BPMN stands for Business Process Model and Notation. It defines specific shapes for tasks, decisions, events, and data flows so that anyone trained in the standard reads the diagram the same way. Most enterprise process tools support BPMN. MemoryLane can export discovered processes as BPMN so you can feed them into existing automation platforms. ### Can I use more than one tool? Yes, and most teams do. The tools in this list solve different problems. MemoryLane discovers which processes exist and how much time they take. Lucidchart or Visio documents them formally. Scribe turns them into step-by-step training guides. They work well together because the output of one feeds the input of another. ### I am ready to buy a team plan. What is the quickest way to get set up with user seats and permissions? Most drawing tools (Visio, Miro, Lucidchart) let you add seats through their admin dashboard. With MemoryLane, [book a call](mailto:founders@trymemorylane.com) and we handle the team setup for you: user seats, privacy controls, and a central dashboard where you see results across the team. Most deployments are live within a week. ### How do I get my team to adopt process mapping? The biggest barrier is not the tool, it is getting people to participate. We covered this in [how to successfully roll out AI](https://trymemorylane.com/blog/how-to-roll-out-ai). Short version: leaders go first, share their own results, and frame it as removing boring work rather than monitoring people. The first automation you build should make someone's week visibly better. ## Next steps Try [MemoryLane](https://trymemorylane.com) on the [self-serve plan](https://trymemorylane.com/#pricing) and read the [guide](https://trymemorylane.com/guide) to get set up in five minutes. If you need help with team deployment, security controls, or a customized analysis, [get in touch](mailto:founders@trymemorylane.com). ================================================================================ # How to successfully roll out AI Source: https://trymemorylane.com/blog/how-to-roll-out-ai ================================================================================ # How to successfully roll out AI **Contents** - [What the best companies do](#what-the-best-companies-do) - [AI champions create momentum](#ai-champions-create-momentum) - [Executives must participate](#executives-must-participate) - [Bottom-up often scales better](#bottom-up-often-scales-better) - [People benefit from automating their own work](#people-benefit-from-automating-their-own-work) - [The playbook](#the-playbook) Every company rolling out AI faces the same choice. Go top-down or go bottom-up. **Top-down** means leadership drives it, buys the tools, and rolls the changes out across the organization. It is fast on paper but can be slow in practice because it may encounter resistance. **Bottom-up** means you start with a few "AI-pilled" people who discover their own automation opportunities, build their own case, and pull others along. Most companies default to top-down because it feels like you don't have to address the "people" side of things. We all know how this ends. ## What the best companies do Top-down AI rollouts follow a predictable script. Leadership picks vendor, runs an implementation project across the organization and rolls the thing out. On paper it makes sense and seems efficient. In practice it often creates resistance within the ranks. And no wonder it does. People feel surveilled and terribly micro-managed. They worry the project exists to replace them or catch them slacking. Even when none of that is true, the perception alone is enough to kill the project. We have seen pilots where employees quietly kept tracking tools turned off for weeks. The technology worked fine, but the trust wasn't there. Ask any CEO what blocks their AI initiatives and you will hear about budget, integration complexity, data quality, and finding the right use cases. These are real issues. But the one that actually kills projects is simpler – most people don't want change. You cannot transform an organization by deciding at the top what changes and pushing it down without getting others onboard. The people doing the work have to want the change. And they will not want it if it feels like something being done to them. The traditional answer is "change management," which usually means a communication plan, a few town halls, and a project manager making sure adoption metrics hit targets and so on. The actual problem is that nobody asked the people doing the work whether they wanted help, what kind of help, or where. ## AI champions create momentum Instead of rolling out AI top-down, the most successful companies we've worked with start with a handful of people who actually want to drive change. Find two or three people who are curious about AI and automation. Maybe they already tinker with tools on their own. Maybe they have complained about repetitive work. Give them access to a [task mining tool](https://trymemorylane.com/blog/task-mining-for-small-teams) and let them run it on their own machines, on their own terms, no centralized data storage, purely individual use. More often than not, something interesting happens. When someone discovers for themselves that they spend six hours a week on a workflow that could be automated, they do not feel threatened. They feel empowered. They see a specific opportunity that feels personal to them. And they become an internal advocate without anyone ever asking them to. This is how you solve the cultural change without ever calling it a problem. You do not need a change management plan because there is nothing to manage. People are choosing to participate. The resistance that kills forced top-down rollouts simply does not exist. ## Executives must participate The fastest way to build trust is for leadership to go first. When an executive runs the tool on their own machine and shares their results with the team, it sends a clear signal. This is not about watching people. There is no hidden agenda. They have skin in the game. "I spend four hours a week on this and we can automate it" lands very differently coming from a CFO showing their own data than from a company-wide mandate that nobody asked for. It removes the fear because it proves there is nothing to fear. If leadership is willing to look at their own workflows, it is hard to argue the tool is there to surveil anyone else. This also gives executives firsthand experience with what the tool actually produces. When they talk to the board about AI adoption, they are not repeating a vendor pitch. They are speaking from their own data. ## Bottom-up often scales better The common objection is that bottom-up adoption is too slow. You need three to five champions with real data and real stories before the approach has any momentum. But those three to five champions are worth more than a 200-page consulting report. They have firsthand experience. They can speak to their colleagues as peers, not as management pushing an agenda. They have specific numbers: "I saved four hours a week on invoice processing" is more convincing than any slide deck. Once you have that proof from the inside, two things can happen. Leadership sees the results and funds a broader rollout with genuine buy-in from the people it affects. Or you skip the top-down step entirely and keep growing bottom-up, team by team. If your organization is large enough that a coordinated top-down rollout makes sense, bottom-up gives you the buy-in to actually pull it off. If you are smaller and a formal rollout feels like overkill, just keep expanding organically. The foundation is real adoption either way. ## People benefit from automating their own work There is also a personal angle that most AI rollout stories ignore. Nobody knows exactly how AI reshapes work over the next five years. But in every scenario, the people who understand their own workflows and can speak concretely about what AI does and does not do well are in a stronger position. If AI augments your job, you will be the one who benefits most because you already know where the leverage is. If AI changes parts of your role, you will be the one who saw it coming and adapted. Understanding your own work patterns is not just useful for your company. It is useful for you. Executives should make this clear to their people. [MemoryLane](https://trymemorylane.com) is how you build that understanding now, on your own terms. It installs in [five minutes](https://trymemorylane.com/guide), runs in the background while you work, and within a week shows you exactly where your time goes and which of that work has automation potential. Your data stays on your device. Screenshots are processed and [immediately deleted](https://trymemorylane.com/security). The [$100/month self-serve plan](https://trymemorylane.com/#pricing) is designed for exactly this: one person exploring their own workflows before involving anyone else. ## The playbook 1. **Executives go first.** Run [MemoryLane](https://trymemorylane.com) on your own machine, share your results with the team. Show there is nothing to fear and that you have skin in the game. 2. **Appoint two or three AI champions.** People who are curious about automation, not people who were told to participate. Let them run the tool on their own terms. 3. **Let champions tell the story for you.** When they share "I found 30 hours a week of copy-paste across our team," that is more convincing than anything leadership can say. Peer proof beats top-down messaging every time. 4. **Expand from proof.** Once you have real results from real people, decide what fits your organization: keep rolling out bottom-up, or use the buy-in to fund a broader top-down initiative. Either path works. ## Next steps Start with yourself. Run [MemoryLane](https://trymemorylane.com) on your own machine, look at your results, and share what you found with your team. Read more about [how to map your processes](https://trymemorylane.com/blog/how-to-map-your-processes) or [why task mining is not employee monitoring](https://trymemorylane.com/blog/task-mining-vs-employee-monitoring) if your team has privacy concerns. [Get in touch](mailto:founders@trymemorylane.com) if you want help with a team rollout. ================================================================================ # Why task mining is not employee monitoring Source: https://trymemorylane.com/blog/task-mining-vs-employee-monitoring ================================================================================ # Why task mining is not employee monitoring **Contents** - [What employee monitoring does](#what-employee-monitoring-does) - [What task mining does](#what-task-mining-does) - [The difference in one sentence](#the-difference-in-one-sentence) - [Why the fear is still legitimate](#why-the-fear-is-still-legitimate) - [How to roll it out without losing the room](#how-to-roll-it-out-without-losing-the-room) Almost every rollout conversation we have includes some version of the same question. "So this records everything on my laptop?" It is a fair question. Task mining works by observing what happens on screen, and on the surface that sounds exactly like employee monitoring software. The two categories get lumped together all the time. They should not be, because they are built for opposite purposes, and confusing them is the fastest way to kill a process discovery project before it starts. ## What employee monitoring does Employee monitoring software exists to answer one question. Is this person working? It tracks individuals. Keystrokes, active and idle time, which websites someone visited, sometimes periodic screenshots that a manager can open and review. The output is a per-person report, often with an "activity score" attached. Managers use it to check on remote workers, catch time theft, or enforce policy. Whatever you think of that (we think it mostly destroys trust and measures the wrong thing), the design goal is clear. The unit of analysis is the person, and the audience is their boss. ## What task mining does Task mining exists to answer a different question. Where does our time actually go, and which of that work could software handle? It observes screen activity too, but it is looking for patterns, not people. Repeated sequences of steps. The same data moved between the same apps every morning. A 12-step workflow that spans five browser tabs and eats hours every week. The output is a map of workflows ranked by time spent and automation potential. Nobody gets an activity score. There is no idle-time report. The interesting result is never "Anna worked 7.2 hours yesterday", it is "invoice processing takes this team 30 hours a week and most of it is copy-paste." The unit of analysis is the process. The audience is whoever wants to fix it. ## The difference in one sentence Employee monitoring evaluates people. Task mining evaluates processes. | | Employee monitoring | Task mining | | ----------------------- | ----------------------------------- | ----------------------------------------- | | **Question it answers** | Is this person working? | Where does time go, what can we automate? | | **Unit of analysis** | The individual | The workflow | | **Typical output** | Activity scores, per-person reports | Ranked list of processes worth automating | | **Who looks at it** | Managers checking on people | Teams improving how work gets done | | **Effect on trust** | Usually corrosive | Neutral to positive, if rolled out right | ## Why the fear is still legitimate Saying "this is not monitoring" is not enough. The tool still sees screens, so the burden is on the software (and on you, if you are rolling it out) to make the distinction real rather than rhetorical. Here is what that looks like in practice, and what we built into [MemoryLane](https://trymemorylane.com): 1. **Data stays on the user's device.** Each person's captured data lives locally. They can choose to share it with the organization for central analysis. It is not silently streamed to a manager's dashboard. 2. **Screenshots are processed and immediately deleted.** The AI reads a screenshot, extracts what workflow step it represents, and the image is gone. Never stored by anyone at any point. We run everything through [zero-data-retention](https://trymemorylane.com/security) AI providers, and on-premise deployment is available if you need tighter controls. 3. **People can see and control what is captured.** Sensitive apps can be excluded. The person being observed should never have to guess what the tool sees. If a task mining tool cannot give you clear answers on those three points, treat it as monitoring software no matter what the marketing says. ## How to roll it out without losing the room The technology only solves half the problem. The other half is how you introduce it. We have seen pilots where employees quietly kept the tracker turned off for weeks, and the project produced almost no data. The tool worked fine. The trust did not. What works instead: 1. **Leaders go first.** Run it on your own machine, share your own results with the team. "I spend four hours a week on this and we can automate it" lands very differently than a company-wide mandate. 2. **Frame it as removing boring work, honestly.** The message is "we are finding the repetitive stuff that wastes your day so we can get rid of it", and the follow-through has to match. The first thing you automate should make someone's week visibly better. 3. **Never use the data for performance conversations.** Not once. The first time workflow data shows up in a performance review, the project is dead and so is the next one. We wrote more about this in [where to start with AI](https://trymemorylane.com/blog/where-to-start-with-ai), which covers finding champions and building adoption from the ground up. ## Next steps If you want to see how your team actually works without turning your company into a surveillance operation, this is exactly what we built MemoryLane for. It installs in [five minutes](https://trymemorylane.com/guide), starts at [$100 a month](https://trymemorylane.com/#pricing), and the [security setup](https://trymemorylane.com/security) is public. [Get in touch](mailto:founders@trymemorylane.com) if you need on-premise deployment, custom data controls, or help getting your team onboard. ================================================================================ # How to map your processes Source: https://trymemorylane.com/blog/how-to-map-your-processes ================================================================================ # How to map your processes **Contents** - [Why most process maps are wrong](#why-most-process-maps-are-wrong) - [Three ways to do it](#three-ways-to-do-it) - [Which approach fits you](#which-approach-fits-you) - [What to do with the map](#what-to-do-with-the-map) What is the best way to map how work gets done? Most companies skip this step entirely. They buy automation tools, pick something to automate based on a gut feeling, and wonder why the impact is underwhelming. The reason is simple. Nobody actually mapped the processes first. Process mapping means creating a clear, accurate picture of how work gets done in your organization. Not the org chart version. Not the job description version. The real version, where someone spends 40 minutes every morning copying data from an email into a spreadsheet and then into an ERP, and nobody talks about it because it has always been that way. ## Why most process maps are wrong The standard approach is to sit people down and ask them to describe their workflows. This produces a process map that looks neat on paper but misses most of what actually happens. People cannot accurately describe work they do on autopilot. The tab switches, the copy-paste sequences, the email follow-ups, the workarounds they invented two years ago. These things happen hundreds of times a week and nobody mentions them because they barely register anymore. We ask every new customer if they have SOPs or process documentation. The answer is almost always no. And when documentation does exist, it describes how work is supposed to happen, not how it actually happens. Those are very different things. ## Three ways to do it ### 1. Manual mapping Interview your team. Watch them work. Draw flowcharts in [Miro](https://miro.com/) or [Lucidchart](https://www.lucidchart.com/) or a spreadsheet. This is the cheapest option and the most common starting point. It works if your team is small (under ten people) and the workflows are straightforward. For anything larger, it almost always stalls. The person running the project gets pulled into other priorities, the flowcharts go stale, and the whole initiative quietly dies. We have seen this happen at companies of every size. The other problem is accuracy. You will capture the main steps of each process, but you will miss the in-between work. The 12-step sequence that spans five browser tabs. The manual data entry that nobody considers part of the process because it only takes "a few minutes" each time, but adds up to hours every week. If you go this route, timebox it aggressively. Give yourself two weeks to map the top five workflows and accept that the map will be incomplete. ### 2. Consultants Hire a firm that specializes in process discovery. At the enterprise level, the [Big Four](https://en.wikipedia.org/wiki/Big_Four_accounting_firms) (Deloitte, EY, PwC, KPMG) all offer this under their "digital transformation" or "intelligent automation" practice. Smaller boutique firms exist too. These people know what they are doing. They have frameworks, they have done it many times, and they deliver a polished report with clear recommendations. The output is better than what most teams can produce on their own. But the cost starts at tens of thousands of dollars and goes up fast. [Gartner estimates](https://www.gartner.com/en/documents/4012770) the average process mining initiative at $250K+ at the enterprise level. Even a smaller engagement with a boutique firm will run five figures minimum. We break down the full range in [how much does process mapping cost](https://trymemorylane.com/blog/how-much-does-process-mapping-cost). The method is still fundamentally interviews and observation. More structured than doing it yourself, but it still depends on people remembering and describing work they do without thinking about it. The other problem is shelf life. A consultant maps your processes in Q3, and by Q4 the map is already wrong. Teams change, tools change, workarounds evolve. A one-time snapshot goes stale fast, and there is no operational monitoring built in to track how work actually flows after the engagement ends. If you have the budget and executive backing, this path delivers. For most small and mid-size teams, it is out of reach. ### 3. Automated discovery This is the newest approach and the one that changes the math for smaller teams. Software observes how work actually happens on people's screens, recognizes the patterns, and maps the processes without anyone having to describe anything. Instead of asking people what they do, you look at what they actually did. Every app switch, every repeated sequence, every chunk of time spent on a specific task. The data is there whether someone thinks to mention it or not. Unlike a one-time consultant engagement, automated discovery runs continuously. It is an operational monitoring tool that tracks how work actually flows in real time, not a snapshot that goes stale after one quarter. When a process changes, the data updates automatically. This approach is called [task mining](https://trymemorylane.com/blog/task-mining-for-small-teams). The enterprise tools that pioneered it (UiPath, Celonis, Mimica) charge six figures and require dedicated implementation teams. We built [MemoryLane](https://trymemorylane.com) to make the same capability accessible to normal-sized companies. MemoryLane installs in [five minutes](https://trymemorylane.com/guide), runs in the background as people work, and within a week produces a ranked list of workflows sorted by time spent and automation potential. No interviews. No flowcharts. No consultants. Your data stays on the device and screenshots are processed and immediately deleted through [zero-data-retention](https://trymemorylane.com/security) AI providers. The [$100/month self-serve plan](https://trymemorylane.com/#pricing) gives a single user full access. When you are ready to roll it out across your team, you upgrade to a team project and run the analysis centrally. ## Which approach fits you **Manual mapping** works for teams under ten with simple workflows and someone willing to own the project for two weeks. **Consultants** work when you have the budget ($50K+), executive sponsorship, and a complex enough operation to justify the spend. **Automated discovery** works when you want accurate results without the time investment of manual mapping or the cost of consultants. It is the only approach that captures the work people do between systems, which is often where the most time goes. Most teams we work with tried manual mapping first, got an incomplete picture, and then switched to automated discovery to fill the gaps. That combination tends to work well. The manual exercise gives you context and buy-in from the team. The automated data gives you accuracy and completeness. ## What to do with the map A process map is not the end goal. The point is to answer two questions: where is time going, and which of those tasks can be handled by AI or automation? Once you have that ranked list, pick the workflow that takes the most time and has the highest automation potential. Build the smallest automation that actually helps. Run it for a week. Measure the result against the baseline your process map gave you. That first win is worth more than the time savings it produces. It gives you proof that the approach works, which is what turns a small experiment into a real initiative with budget and support behind it. ## Next steps Start by getting visibility into how your team actually works. Try [MemoryLane](https://trymemorylane.com) on the [self-serve plan](https://trymemorylane.com/#pricing), or read more about [where to start with AI](https://trymemorylane.com/blog/where-to-start-with-ai) if you are still figuring out your approach. [Get in touch](mailto:founders@trymemorylane.com) if you need help with team deployment, security controls, or a customized analysis. ================================================================================ # Celonis vs UiPath vs Mimica: Task Mining Compared (2026) Source: https://trymemorylane.com/blog/memorylane-vs-uipath-celonis-mimica ================================================================================ # Celonis vs UiPath vs Mimica: Task Mining Compared _Disclosure: we make [MemoryLane](https://trymemorylane.com), the fourth tool in this comparison. Every tool here gets an honest breakdown._ **Contents** - [Celonis](#celonis) - [UiPath](#uipath) - [Mimica](#mimica) - [MemoryLane](#memorylane) - [How they compare](#how-they-compare) The enterprise process intelligence market has some excellent tools. [Celonis](https://www.celonis.com/), [UiPath](https://www.uipath.com/), and [Mimica](https://www.mimica.ai/) all built products that help large organizations understand and improve how work gets done. But there is a gap. These tools are built for companies with hundreds or thousands of employees, dedicated process teams, and budgets starting in the six figures. That leaves roughly 90% of companies without an option. This post breaks down what each tool does well, where it falls short, and who it is built for. ## Celonis Celonis is the leader in process mining. Process mining is a process discovery approach that works by plugging into enterprise systems (think SAP, Oracle, Salesforce) and analyzing the "event logs" those systems produce. In other words, connect your SAP instance and Celonis tells you how long the average invoice approval takes across 10,000 cases. It finds bottlenecks, compliance violations, and inefficiencies at massive scale. The catch is that process mining needs structured data to work with. Someone has to configure the connectors, clean and transform the system logs into a readable format, and build the dashboards. That is typically a team of analysts working for weeks or months. That's why Celonis deployments are large enterprise projects with price tags to match. Process mining also only sees what happens inside the systems it connects to. It does not see the work that happens between systems. The copy-paste from the ERP to the spreadsheet, the manual email follow-up, the 12-step workflow that spans five browser tabs. For smaller teams, that in-between work is where most of the time actually goes. Process mining misses it entirely. ## UiPath UiPath was the first major player in RPA (Robotic Process Automation) and has grown into one of the biggest names in automation. They are publicly traded with over $1.4 billion in annual revenue. Their software covers everything from process mining to RPA bots and AI-powered document understanding. All of the above makes the product powerful, but deploying it often requires something called a "center of excellence," a dedicated internal team that manages the platform, builds the automations, and runs the process discovery. That team alone can cost more than what most mid-size companies spend on their entire finance function. UiPath is built for organizations with thousands of employees and complex IT infrastructure. If you run a 50-person company or a finance team of five, UiPath is not built for you. Not because the product is bad, it's just that the model isn't made for smaller companies. ## Mimica Mimica is the closest to what we do. They observe actual screen-level activity instead of relying on system logs. That means they can see the real work, not just what an ERP captures. But Mimica still targets large enterprises. Their focus is companies with 1,000+ employees. Implementation involves working with their team of analysts to scope the project, deploy the software, and interpret the results. It is a lighter engagement than UiPath or Celonis, but it is still a managed project rather than something you can start on your own. If you run a mid-size company or a small team, Mimica is unlikely to take your call. And if they do, the price tag won't be what you expect. They are a great product solving a painful problem for a different market segment. ## MemoryLane We built MemoryLane for the companies these tools leave behind. You do not need a center of excellence. You do not need to connect your SAP instance. You do not need an analyst team to interpret the output for you. You install it in [five minutes](https://trymemorylane.com/guide), it runs in the background as you work, and within a week you have a ranked list of workflows sorted by time spent and automation potential. The [$100/month self-serve plan](https://trymemorylane.com/#pricing) gives a single user full access to AI-driven process discovery and analysis. When you are ready to roll it out across your team, you upgrade to a team project and run the analysis centrally. No six-month implementation timeline. You can be live in a few days. We've also put a lot of work into data privacy. For all users involved, all captured data stays on their device. They can then choose to share it with your organization centrally, so it can be analyzed and actioned. The screenshots we use to track repetitive tasks are processed and immediately deleted, never stored by anyone at any point. We do this by running through [zero-data-retention](https://trymemorylane.com/security) AI providers. And if you need tighter controls, on-premise deployment is available. The goal is the same – see how work actually happens and find what to automate first. The difference is that MemoryLane makes process discovery and [task mining](https://trymemorylane.com/blog/task-mining-for-small-teams) accessible to normal people, not just companies with a process excellence department. ## How they compare | | Celonis | UiPath | Mimica | MemoryLane | | -------------------------- | --------------------------------- | ------------------------- | ----------------------------- | ------------------------------------------------ | | **Target** | 2,000+ employees | 5,000+ employees | 1,000+ employees | Individuals and SMBs | | **Approach** | System log analysis (SAP, Oracle) | RPA + process mining | Screen-level observation | Screen-level summarization | | **Setup** | Analyst team, weeks to months | Center of excellence team | Managed project with analysts | 5 minutes, self-serve | | **Sees work between apps** | No | Limited | Yes | Yes | | **Budget** | $200K+ | $250K+ | $100K+ | [$100/month](https://trymemorylane.com/#pricing) | We wrote more about this gap in [task mining for small and mid-size teams](https://trymemorylane.com/blog/task-mining-for-small-teams) and about [where to start with AI](https://trymemorylane.com/blog/where-to-start-with-ai) if you are still figuring out your approach. ## Next steps If you are evaluating process discovery tools and your company is not a Fortune 500, start with [MemoryLane](https://trymemorylane.com). Try the [self-serve plan](https://trymemorylane.com/#pricing) and [get in touch](mailto:founders@trymemorylane.com) if you need help with team deployment, security controls, or a customized analysis. ================================================================================ # Task mining for small and mid-size teams Source: https://trymemorylane.com/blog/task-mining-for-small-teams ================================================================================ # Task mining for small teams **Contents** - [Enterprise task mining](#enterprise-task-mining) - [What smaller teams need](#what-smaller-teams-need) - [How MemoryLane fills the gap](#how-memorylane-fills-the-gap) - [The visibility gap](#the-visibility-gap) Task mining means observing how work actually happens on people's screens and letting software find the patterns, instead of asking people to describe what they do all day. Until recently, the only tools that could do this were built for enterprises with six-figure budgets and dedicated IT teams. And the only alternative was human consultants. But if you run a 50-person company, a growing e-commerce brand, or a finance function with five people juggling spreadsheets and ERPs, you had two choices. Map your processes by hand (which rarely works, as we wrote about in [Where to start with AI](https://trymemorylane.com/blog/where-to-start-with-ai)), or skip the visibility step entirely and hope for the best. Most teams hope for the best. ## Enterprise task mining The big players in this space are [UiPath](https://www.uipath.com/product/process-mining), [Celonis](https://www.celonis.com/), and [Mimica](https://www.mimica.ai/). They built really good products. But they are built for organizations with hundreds or thousands of employees, dedicated process excellence teams, and budgets to match. This is why a typical enterprise deployment involves weeks of scoping, integration with your existing IT systems, custom dashboards, and ongoing analyst support. [Gartner puts](https://www.gartner.com/en/documents/4012770) the average cost of a process mining initiative at $250K+. That number includes the software, the consulting, and the time your internal team spends managing the project. For a company with 5,000 employees processing millions of transactions, that investment makes sense. But for a team of 10 or 50 people, the math does not work. ## What smaller teams need Smaller teams do not need custom dashboards or enterprise integrations. They need answers to two simple questions. What are we spending time on? And which of those things can AI or automation handle? You need a tool that observes how work happens for a week and gives you a ranked list of workflows sorted by time spent and automation potential. This workflow takes 12 hours a week. This one takes 8. Here is which ones AI can handle today, and here is the estimated annual cost of each. Now, that's something to work with. ## How MemoryLane fills the gap We built [MemoryLane](https://trymemorylane.com) because we spotted this gap. Enterprise task mining worked, but nobody was making it accessible to smaller teams. MemoryLane installs in [about five minutes](https://trymemorylane.com/guide), runs securely and privately in the background as you work, and produces a prioritized list of what you can automate. Your data stays on the device. Screenshots are processed and immediately deleted. [Zero-data-retention endpoints](https://trymemorylane.com/security) and on-premise deployment is available if you need tighter controls or a centralized rollout. The best part is that it starts at [$100 a month](https://trymemorylane.com/#pricing). And because you get useful results within the first week, it's a no-brainer in terms of value for money. ## The visibility gap We talk to a lot of teams that have already bought automation tools. They have [n8n](https://n8n.io/) or [Make](https://www.make.com/) or [Zapier](https://zapier.com/) set up. They have access to [Claude](https://claude.ai/) or Copilot. The tools are not the bottleneck. The bottleneck is knowing where to point them. Without visibility, you automate whatever someone complained about last meeting. The workflow that quietly eats 15 hours a week goes untouched because nobody thought to mention it. This is why task mining exits to close that gap. ## Next steps Start by getting visibility into how your team actually works. Try [MemoryLane](https://trymemorylane.com) on the [self-serve plan](https://trymemorylane.com/#pricing) and [get in touch](mailto:founders@trymemorylane.com) if you need help with a customized rollout, security and data controls, analysis, or automation. ================================================================================ # Where to start with AI Source: https://trymemorylane.com/blog/where-to-start-with-ai ================================================================================ # Where to start with AI **Contents** - [See the work first](#see-the-work-first) - [Option 1: Do it yourself](#option-1-do-it-yourself) - [Option 2: Consultants](#option-2-consultants) - [Option 3: Automated discovery](#option-3-automated-discovery) - [Getting your team onboard](#getting-your-team-onboard) - [Start small](#start-small) Every company we talk to has the same story. They bought the automation tools and then did not know what to do with them. They did not know where to start. That is completely normal. Whether you are vibe-coding your automations with LLMs or running them on [n8n](https://n8n.io/), [Make](https://www.make.com/), [Zapier](https://zapier.com/), or even just [Claude](https://claude.ai/), you likely have a hard ceiling on how much impact you can make. The reason is simple. You do not know how your company actually runs day to day. You do not have the visibility you need into the real work. That is where you need to start to make any of this work. We have seen this play out dozens of times. A team picks something flashy to automate, it saves twenty minutes a month, and everyone quietly loses interest. The fix is boring but it works. You have to actually look at the work before you try to improve it. ## See the work first We ask every new customer the same question at kickoff: do you have any SOPs (Standard Operating Procedures) or process notes for the work you want to improve? The answer is almost always no. Nothing is written down. The processes live entirely in people's heads. And you cannot automate what you cannot see. Most companies have job descriptions and org charts and maybe a few SOPs that nobody has updated since 2022. None of that tells you which tasks eat three hours a day or which workflows involve seventeen tab switches between five different apps. You need an accurate picture of the real work before you can make any good decisions about what to automate, what to simplify, and what to leave alone. There are three ways to get that picture, and they are very different in cost and reliability. ## Option 1: Do it yourself Sit down with your team. Interview people. Watch them work. Write the processes down in a document or draw them in something like [Miro](https://miro.com/) or [Lucidchart](https://www.lucidchart.com/). This rarely works well. People cannot accurately describe work they do on autopilot. The invoice downloads, the status update emails, the copy-paste between the ERP and the spreadsheet. These tasks happen hundreds of times a week and nobody mentions them in an interview because they barely notice them anymore. Manual process mapping is also brutally time intensive. A single department can take weeks. And the initiative almost always loses momentum before it finishes. The person running it gets pulled into other priorities, the flowcharts go stale, and the whole effort quietly dies. We have watched this happen at companies of every size. If your team is five people with straightforward workflows, it can work. For anything bigger, the effort almost never pays off. ## Option 2: Consultants Hire outside help. There are plenty of firms that specialize in process mapping and operational efficiency. At the enterprise level, the [Big Four](https://en.wikipedia.org/wiki/Big_Four_accounting_firms) (Deloitte, EY, PwC, KPMG) all offer this, often under their "intelligent automation" or "digital transformation" practice. These people know how to run process discovery projects. They have frameworks, they have done it many times, and they deliver a polished report with clear findings. But it is expensive. A meaningful engagement starts at tens of thousands of dollars and goes up fast. [Gartner estimates](https://www.gartner.com/en/documents/4012770) that the average process mining initiative costs $250K+ at the enterprise level. And even with experienced consultants, the method is still mostly interviews and observation. More structured than doing it yourself, but it still depends on people remembering work they do without thinking about it. If you have the budget and leadership backing, this path delivers. For small and mid-size teams, it is simply out of reach. ## Option 3: Automated discovery Five years ago this option did not exist. AI has gotten good enough to observe how work happens on a screen, recognize the patterns, and map them without anyone having to describe anything. We built [MemoryLane](https://trymemorylane.com) to do exactly this. It runs in the background on your team's machines, observes the real pattern of apps, windows, and repeated steps over a normal work week, and produces a ranked list of the workflows worth automating first. No interviews. No flowcharts. No six-figure consulting bill. You are not relying on what people remember. You are looking at what actually happened. Every tab switch, every repeated sequence, every chunk of time spent in a specific app. The data is there whether someone thinks to mention it in an interview or not. Setup takes about [five minutes per machine](https://trymemorylane.com/guide). Your data stays on the device. Nothing leaves unless you choose to share it, and even then, everything runs through [zero-data-retention endpoints](https://trymemorylane.com/security). Within a week you have a picture of where time goes that would take months to build any other way. The enterprise process mining tools like [UiPath](https://www.uipath.com/product/process-mining), [Celonis](https://www.celonis.com/), and [Mimica](https://www.mimica.ai/) do similar things, but they are priced for large organizations and require significant integration work. We wrote more about this gap in [task mining for small and mid-size teams](https://trymemorylane.com/blog/task-mining-for-small-teams). MemoryLane starts at [$100 a month](https://trymemorylane.com/#pricing) and works out of the box. ## Getting your team onboard If you want to push AI adoption in your company, you need people onboard. The best way to do that is to lead by example. Start with yourself. Run the tool on your own machine, look at the results, and share what you found with the team. When an executive or team lead shows their own data and says "look, I spend four hours a week on this and we can automate it," it removes the fear. People see a concrete example of the value instead of a vague pitch about AI. Then find two or three champions. People who are already excited about automation and want to try it. Give them access and let them run it. When they showcase how they benefited, it brings others along naturally. We see significantly better adoption when it spreads this way compared to a top-down mandate that nobody asked for. Frame the rollout as the company letting AI handle the boring manual work so people can focus on work worth their time. That message matters. Nobody wants to hear "we are monitoring your workflows." They want to hear "we are finding the stuff that wastes your day so we can get rid of it." ## Start small Once you have the ranked list of workflows, pick one. Just one. Whichever tool gave you the visibility already tells you how long it takes and how often it happens. That is your baseline. Build the smallest automation that actually helps. One step taken off someone's plate, done reliably. Run it for a week and compare against that baseline. If it holds up, you have something better than a time saving. You have proof that the approach works. And proof is what turns a single win into a budget, a roadmap, and a team that actually wants to do the next one. ## Next steps Start by getting visibility into how work actually happens. Try [MemoryLane](https://trymemorylane.com) on the [self-serve plan](https://trymemorylane.com/#pricing) and [get in touch](mailto:founders@trymemorylane.com) if you need help with a more customized rollout, security and data controls, analysis, or automation.