memrecord
Your entire workspace memory. Processed locally. Powered by on-device AI.
An instant, searchable timeline on your laptop that unifies scattered client emails, WhatsApp logs, contracts, and project files. Automatically indexed and queryable via local AI. 100% offline, zero cloud uploads, completely private.
From years of context down to a single message
A timeline built automatically on your machine by local AI. Scroll through years of client history and zoom into any level of granularity: from high-level project milestones down to the exact email or WhatsApp chat where it happened.
Zoom out to quarters. Zoom in to the exact minute.
Clients
Corporate
Acme Corp
2026 · Q1
Jan — Mar12 items
- Phase 3: Handover and maintenance contract
2025 · Q4
Oct — Dec89 items
- Phase 2: Frontend implementation & API auth
- Major blocker: Auth0 integration delays
2025 · Q3
July — Sept34 items
- Phase 1: Database architecture & design
- Total invoiced: $12,500
Oct 12Scope change
Client requested adding Google OAuth to the login flow. Agreed to extend the deadline by 2 weeks.
Oct 18Invoice sent
Billed $4,000 for halfway milestone. Paid on Oct 20.
Oct 22Technical blocker
Client's AWS server went down. We paused development.
Email from: Dave (Acme)
Re: Auth scope
Hey, the board just met. We really need Google OAuth added to the login screen.
WhatsApp to: Dave
We can add Google OAuth, but it's out of the original scope. I'll need an extra 2 weeks on the deadline to implement it.
10:30
WhatsApp from: Dave
That's totally fine. Take the extra 2 weeks. Just make sure it works flawlessly.
10:45
Interactive relational graph
Navigate your workspace through connected relationships instead of flat folders. Scroll down any branch to automatically filter files and give your AI precise context as you search.
Semantic search, not rigid keyword matching
Unlike Spotlight, which fails unless you remember the exact filename or phrase, local AI understands context and meaning. Ask natural questions about your work history to retrieve exact messages, scope changes, and invoices without needing exact keyword matches.
Ask local AI
Client: Acme Corp
Turn scattered files into a unified vault
Simply bring your existing chat histories, calendar files, invoices, and developer logs. Local AI handles the background parsing, transcription, and structuring 100% on-device.
Your secured, local-only vault of information
▸ parsing whatsapp export · 1,204 messages
- Indexed
- 4,812
- Entities
- 37
- Vault Size
- 4.81 GB
Google Calendar
.ics
.txt / .zip
Skype
.tar
Slack
.json
Gmail
.mbox / .eml
Obsidian
.md
Invoices
Git
git log
Your data is not training fuel
Cloud AI tools treat your client contracts, code repositories, and chat logs as fuel for their future commercial models. This vault keeps them locked entirely on your device.
01. On-Device Execution: Search indexing, vector embedding, and AI response generation run 100% inside your local hardware.
02. Strict NDA Integrity: Zero cloud sub-processors or third-party APIs means total compliance with client confidentiality agreements.
03. Encrypted at Rest: Your local index is fully encrypted on your physical drive using SQLCipher protocols.
Local hardware sandbox
Secure your workspace memory
Download the native desktop client and index your workspace locally. No cloud tracking and zero API keys required.
Hardware & protocol specs
- Q: What models run on my machine?
- A: The desktop app profiles your laptop hardware on launch. Standard setups (8GB RAM) run highly optimized 3B parameter models. Pro setups (16GB+ unified memory/RAM) automatically unlock full 8B parameter models for deep context reasoning across years of history.
- Q: Can I view my timeline when I'm away from my laptop?
- A: Yes. The desktop app includes a lightweight mobile companion viewer. It mirrors your local SQLite index directly to your phone over your local Wi-Fi network using a zero-cloud, peer-to-peer TLS handshake.
- Q: Will it drain my laptop battery?
- A: No. File ingestion and vector indexing run as a low-priority background task only when new documents are added. Reading your timeline and querying the local AI exert near-zero overhead on system resources.