Cognitive memory engine for AI agents. Sub-millisecond recall using Sparse Distributed Representations (SDR) instead of embeddings — no LLM calls, no cloud, no external dependencies. Pure Rust core with Python bindings, ~3MB binary, ChaCha20+Argon2id encryption at rest. Hierarchical memory with biological-inspired decay, trust and provenance scoring, PII detection. Integrates with LangChain, CrewAI, AutoGen, OpenAI Agents, LlamaIndex, and Claude MCP. Patent pending US 63/969,703.
Fund this project