RushDB

Pairing vector search with the graph layer is the interesting part for retrieval - agents can jump from a semantic match to connected context in one hop. Is the vector index built in, or do you bring your own embeddings?

The no-schema "push any JSON, get relationship-aware records" is the part that sells it — schema migrations are exactly what kills using a real graph DB as agent memory. The thing I'd want to know before wiring it into an agent loop: across many writes of the same entity in slightly different shapes, does it dedup/merge into one node, or does each push create new ones? Unbounded node growth is what quietly turns agent memory into agent landfill. Built on Neo4j is reassuring either way — starring it.