Engram Persistent Memory

Example Use Cases

• 1

  Persistent, structured episodic and semantic memory storage.

• 2

  That require LLM-powered reasoning across multiple memory types.

• 3

  Federated search across external knowledge providers.

Description

Engram-Mem is a dual-memory AI system combining episodic vector memory and semantic graph memory with large language model reasoning. It supports entity-gated ingestion to store only meaningful data and offers multi-tenancy, authentication, caching, and observability features. The server integrates with various AI agents and IDEs, federates with external knowledge systems, and exposes multiple interfaces including CLI, MCP stdio, HTTP API, and WebSocket for flexible usage.