Vector-symbolic architectures, hyperdimensional computing, memory systems, and agentic runtimes. Every project we publish is real software under a real license — no papers without repos.
Ada — a schema-on-write memory system for LLMs. Every fact is mapped into a universal 7-layer × 33-role lattice at write time, so recall is a deterministic scan: exact answers, versioned history, no confabulation, and no LLM in the read path. Speaks MCP, runs on SQLite or Postgres.
View on GitHub →The open catalog of pure-HDC models for the Glyphh runtime — classification, churn prediction, intent parsing, prompt-injection detection, and more. Character-level encoding, morphology, and syntax with no LLM in the loop: inference in microseconds on a CPU.
View on GitHub →The Ada research program — the theory line behind adaL. Universal fact schemas, deterministic recall, versioned belief, and where vector-symbolic representations meet structured memory.
Follow the org →The agentic runtime everything else plugs into. Hosts HDC models from the catalog and MCP tools next to your agents, giving them state, memory, and compute over open protocols.
Follow the org →Agentic coding on the Glyphh runtime — an agent with real memory via Ada and HDC-fast intent routing, working in your repo. Also ships as a product.
See the product →github.com/glyphh-aiEverything lives here — releases, roadmaps, and the projects that haven't been announced yet.
Visit the org →New projects, models, and releases land on GitHub first. Star the org, read the research, or run the products.