Persistent AI memory
Local-first code-graph memory layer for AI coding agents. Symbol resolvers for Rust, TypeScript, Python. Symbol-anchored embeddings. Shipping today as Mneme v0.4.0.
//02Six research areas
Structural certainty at scale
Real symbol resolvers across more languages. Cross-repo call graphs that survive renames. Goal: 10/10 canonical recall on the golden benchmark.
Local-first LLM routing
When should an AI host route to a local 3B/7B model vs a cloud frontier model? Latency, cost, privacy — all inputs. We're building the runtime decision engine.
Cross-project federation
Code-graph signatures hashed and shared between machines you control. Lets your AI answer "have we seen this pattern before?" — without ever sending source code anywhere.
Multimodal recall
Architecture diagrams, screenshots, voice notes — all indexed and recallable alongside code. All stitched into the same memory graph.
AI-native IDE primitives
What does an editor look like when AI is a first-class collaborator with persistent memory and structural understanding? Not a sidebar — a co-author with its own workspace.
//03Open questions
Six open questions, if you want to talk.
- What does persistent memory look like for an AI agent that lives a hundred sessions long?
- Can structural certainty fully replace text matching, or do we always need the fallback?
- How do you make federation private enough that engineers actually turn it on?
- What's the right boundary between local-first compute and cloud-first models?
- Is the right unit of AI memory a file, a symbol, or a decision?
- When the AI co-author has its own memory, who owns the institutional knowledge?