Semantic memory (pgvector) #19

Closed
opened 2026-01-28 19:04:53 +00:00 by jason.woltje · 1 comment
Owner

Implement semantic memory using pgvector.

Features:

  • Store embeddings for tasks, events, notes
  • Semantic search (find similar items)
  • Context retrieval for AI queries
  • Memory consolidation

API:

  • POST /api/memory/embed - Store embedding
  • GET /api/memory/search - Semantic search
  • GET /api/memory/context - Get context for AI

Implementation:

  • Generate embeddings via Ollama (nomic-embed-text)
  • Store in memory_embeddings table
  • Vector similarity search

Requirements:

  • Automatic embedding on create/update
  • Background processing for batch embedding
  • Efficient nearest-neighbor search
Implement semantic memory using pgvector. **Features:** - Store embeddings for tasks, events, notes - Semantic search (find similar items) - Context retrieval for AI queries - Memory consolidation **API:** - POST /api/memory/embed - Store embedding - GET /api/memory/search - Semantic search - GET /api/memory/context - Get context for AI **Implementation:** - Generate embeddings via Ollama (nomic-embed-text) - Store in memory_embeddings table - Vector similarity search **Requirements:** - Automatic embedding on create/update - Background processing for batch embedding - Efficient nearest-neighbor search
jason.woltje added this to the M3-Features (0.0.3) milestone 2026-01-28 19:04:53 +00:00
jason.woltje added the p1aiapi labels 2026-01-28 19:04:53 +00:00
Author
Owner

Absorbed by Knowledge Module.

See: #68 (pgvector Setup), #69 (Embedding Pipeline), #70 (Semantic Search API)

Milestone: Knowledge Module (0.0.5)

Absorbed by **Knowledge Module**. See: #68 (pgvector Setup), #69 (Embedding Pipeline), #70 (Semantic Search API) Milestone: Knowledge Module (0.0.5)
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: mosaic/stack#19