Skip to main content

Roadmap

This roadmap is adapted from the current product and architecture notes, grouped into readable phases.

V2: Deeper Retrieval And Operations​

  • Vector database integrations beyond Postgres, including Pinecone, Weaviate, and Qdrant
  • Memory analytics and operator dashboards
  • Real-time sync and event streaming improvements
  • Richer webhook tooling and delivery controls

V3: Multi-Modal And Cross-System Memory​

  • Multi-modal memory for images, audio, and video
  • Memory compression and summarization for older episodes
  • Better lifecycle tooling for large memory corpora
  • Stronger admin and compliance workflows

V4: Platform-Scale Memory​

  • Federated search across controlled boundaries
  • Alternative query surfaces such as GraphQL
  • More advanced governance and policy features
  • Broader ecosystem integrations for enterprise deployment patterns

What V1 already focuses on​

  • Durable episode storage
  • Hybrid retrieval
  • Session-aware short-term memory
  • Soft deletes and restore
  • Eight framework adapters
  • Self-hosted Docker-first deployment