Memorysystemsinspiredbyhowhumansremember
We're building AI memory systems and applications that understand meaning, form associations, and retrieve through context—not keyword matching.
Our tools take inspiration from human memory principles and Hopfield networks, where memories are stable states in an energy landscape.
What you can build
Agent Memory
Give AI agents persistent, semantic memory that improves with every interaction
Personal Knowledge Base
Build a second brain that connects ideas and surfaces insights through association
Context-Aware Search
Retrieve memories with connected context using graph traversal
Research Tools
Track learnings, connect findings, and build knowledge incrementally
How it works
Hybrid retrieval combining semantic embeddings, fuzzy matching, and lexical search. Finds what you mean, not just what you typed.
Bidirectional connections between memories. Graph traversal for context. Information retrieves related information automatically.
Single-file format that works across systems. Designed for AI agents and human collaboration.