Ori Mnemos Memory Infrastructure

Example Use Cases

• 1

  Persistent, cross-session memory with cognitive forgetting and learning.

• 2

  Efficient multi-hop and recursive retrieval over a knowledge graph.

• 3

  That require zero cloud dependency and full data ownership with local storage.

Description

Ori Mnemos implements human cognition models on a knowledge graph to provide persistent memory across sessions, clients, and machines without cloud dependencies. It features recursive graph traversal, spreading activation, Hebbian learning, and reinforcement learning to optimize retrieval quality over time. The system stores data as markdown files on disk with git version control, enabling zero-infrastructure retrieval that outperforms existing solutions in benchmarks.