Manex

Manex is a browser-first, local-first AI memory tool that helps teams stop re-answering the same document questions. It indexes private documents, generates source-backed answers with citations, and preserves corrections and decisions as reusable team memory. Unlike generic chatbots, Manex tracks contributors, handles stale context with tombstones, and shares derived memory (chunks and embeddings) without redistributing original files.

• Local-first workspace — Private documents and embeddings stay in the browser by default; nothing is bulk-crawled or sent to the cloud unless you choose to sync.
• Grounded answers with evidence — Every answer cites the document chunks, memory corrections, and team contributors that backed it, so you can verify confidence before acting.
• Corrected answer memory — When a teammate fixes an interpretation, the correction is saved as reusable context that shapes future answers instead of disappearing in chat history.
• Stale decision handling — Tombstones and updated corrections override outdated memory so the team consistently gets the current answer, not the old one.
• Team Brain workspace — The workspace owner creates a shared Team Brain and invites members via invite code; only the creator pays, members join for free.
• Encrypted snapshot sync — Shared workspace snapshots are encrypted at the app layer before storage or sync, then transported over HTTPS.
• Agentic file connectors — Ask for a file by intent rather than folder path; Manex searches Google Drive or OneDrive, presents candidates for user approval, then imports only selected files.
• Contributor attribution — Answers show which team member or brain component contributed the evidence, supporting accountability and audit trails.
• Compliance, consulting, and research use cases — Designed for teams where bad reuse creates real cost: inconsistent client answers, slow onboarding, repeated research, or stale policy interpretation.