mihaibuilds

@rivestack No edge weights yet — just typed edges. Good call though, weighting by frequency or recency would make graph recall sharper. WITH RECURSIVE is the plan for traversal. Keeping it simple for v1, graph features expand in milestone 7.

@mihaibuilds that adjacency list pattern is solid in postgres. WITH RECURSIVE handles traversal really cleanly if you need to expand neighborhoods. are you storing edge weights? makes recall across the graph a lot more targeted

I got tired of re-explaining myself to Claude every session. So I built a local-first AI memory system and I'm open-sourcing it. PostgreSQL + pgvector, hybrid search, MCP, knowledge graph. Self-hosted, MIT licensed. 🔗 github.com/MihaiBuilds/me… #opensource #buildinpublic #AI

@rivestack Adjacency in Postgres — entities + relationships tables alongside the vector columns. Standard adjacency list, simple JOINs. One DB, one backup, one connection string. Architecture doc: github.com/MihaiBuilds/me…

@mihaibuilds yeah happy to compare notes. curious how you're handling the graph layer, adjacency in postgres or a separate structure alongside the vector columns?

@YearView Appreciate that. Local-first was non-negotiable — your AI memory should stay on your machine. More coming as the build continues.

@mihaibuilds This is the kind of problem-solving that makes open source great — you hit a real pain point and just... built the solution. Local-first is such a smart call too. Respect for shipping it.

@rivestack Thanks! PostgreSQL + pgvector keeps the stack simple — one database for vectors, full-text, and relational data. No need for a separate vector DB. Would be happy to compare schema approaches as the project progresses.

@mihaibuilds solid approach. postgresql + pgvector is underrated for this kind of persistent memory layer. the hybrid search on top of the graph is what actually makes recall usable. we use the same combo at rivestack.io if you want to compare notes on the schema side