Feather DB Docs.
The Living Context Engine — context infrastructure for AI agents. Embedded, open source, sub-millisecond. Start with a pip install and ship.
Core capabilities
Adaptive Memory
Stickiness × decay × importance. Memory that ages like you do.
Context Graph
Typed edges, BFS traversal, bidirectional index. Knowledge that connects.
Semantic Search
HNSW with AVX2/AVX512 kernels. <1ms at 100K vectors, C++ core.
Hybrid BM25 + Vector
Reciprocal Rank Fusion of dense + sparse — same RRF as Qdrant, Weaviate, Elastic.
Atlas Admin SPA
Brand-aligned dashboard at /admin/ — records, search, schema, hierarchy, graph, context, console. New in v0.10.
Pluggable Embeddings
Six providers (OpenAI, Azure, Gemini, Voyage, Cohere, Ollama). One-call ingest_text — embed + store. New in v0.10.
Metadata Filters
Namespace, entity, attribute, importance — all inside the core engine.
Next steps
Need help?
Open an issue on GitHub, join the Discord, or book 30 min with the founder.