Memory that reasons. Answers that trace.
We're building the knowledge-graph memory layer for your AI coding assistant — structured, traversable, and grounded in your real code.
Why we started
Coding assistants kept failing in the same two ways: they couldn't reason over how a codebase connects, and they couldn't show where their answers came from. Grep and vector search made the first problem worse and ignored the second. So we built Graphify — a knowledge graph your assistant traverses instead of grepping, with every relationship tagged by confidence.
What we believe
Memory should be structured, current, and on-device. And in a world where assistants write and ship code, "looks right" isn't good enough — answers should trace to a real path through your codebase. That thesis runs through everything we build.
Grounded over promises
If we claim it, you can run the command and check it.
Structure over similarity
Meaning lives in relationships, not cosine distance.
Open by default
The core is open source; trust is earned in the open.
Team
Plus a small founding team — we're hiring →