Memory that your team can operate, not just deploy.
Most memory tools optimize for retrieval quality. MemexAI optimizes for operability — memory that is scoped per user, correctable by your team, and auditable when something goes wrong.
The difference is what happens when memory is wrong.
Retrieval-first memory is hard to debug when it fails — wrong extraction, retrieval miss, or retrieval ignored. MemexAI separates those failure modes by keeping memory as auditable files with full revision history and access logs.
Memory should be correctable
Wrong facts should be editable by your team — not buried in a retrieval index. MemexAI stores memory as files your ops team can open, fix, and verify.
Memory should be auditable
Every read and write is logged with a tool call ID. Every file change creates a revision. You can trace any behavior back to the exact memory state that caused it.
Memory should be scoped
Per-user memory is physically isolated. Shared guidance is separate from user facts. Agents never cross tenant boundaries by design.
Fits where memory is a product requirement, not an afterthought.
From personal assistants to multi-agent pipelines — wherever users or your team need to trust, inspect, or correct what the AI remembers.
Per-user memory that your ops team can inspect and fix
Per-user memory scoped and isolated by design, with a correction surface your ops and support teams can actually use.
Read moreMemory that grows with the user and stays correctable
Memory that feels earned — inspectable files the user or your team can open, correct, and trust over time.
Read moreMemory that carries context across tickets and sessions
Carry context across every ticket so users never have to repeat themselves again.
Read moreTask state that survives across context resets
Task checkpoints and learned constraints that survive context resets so agents continue rather than restart.
Read moreShared behavioral context across agents in a product
One shared memory layer that all agents in your deployment read — update once, all agents reflect it.
Read moreMemory you can debug without guessing what the agent learned
Access logs, revision history, and per-file inspection that separate write failures from retrieval failures.
Read moreYour sales AI should remember the last call. And the one before.
Deal context, objection history, and stakeholder map persist across every call — and sales ops can correct stale facts before they cost a deal.
Read moreKeep agent memory clean and accurate over time
As agents write to memory, files grow noisy and redundant. Run a scheduled compaction pass to merge duplicates, remove stale facts, and keep the prompt block tight.
Read moreExtract structured memory from conversation transcripts
Bootstrap agent memory from existing session logs. Turn months of conversation history into structured, inspectable memory files without replaying every message live.
Read moreAn AI tutor that actually remembers how this student learns.
Learner profiles that grow with every session — correctable by teachers and parents, never locked in an opaque model.
Read morePick the integration path that matches your stack.
MemexAI works in two modes that share the same Postgres schema and tool API. Choose direct Postgres for full control, or the Docker service for multi-tenant isolation with an admin UI out of the box.
Direct Postgres
Your app owns the database connection. No HTTP layer. Import @memexai/core, call migrate() once, and wire memory tools directly into your model call.
Read the docsDocker service
Run the MemexAI service alongside your app. Clients use @memexai/sdk with an API key. Postgres, the service, and the admin UI all start from one compose file.
Read the docsAgent onboarding
Hand a coding agent the setup file at memexai.space/setup.md and it wires MemexAI into your existing agent project — detecting your stack and adding the right adapter.
Read the docsAdapters for the frameworks your agents already use.
MemexAI ships adapters for Vercel AI SDK, Anthropic SDK, LangChain, OpenAI, LlamaIndex, CrewAI, and Google ADK — in TypeScript and Python. You keep your existing model SDK and add memory at the agent boundary.