About Oynix
Nothing your team learns should be lost.
AI coding agents are capable and blind. They understand code. They do not understand organizations. They do not know why something exists, who owns it, what a thread decided, or which choice still matters. Every conversation starts from zero. Oynix fixes that.
Oynix is the memory layer for AI coding agents. It ties your code, tickets, docs and conversations into one live knowledge graph, then serves that graph to whatever agent you use over MCP. The agent stops re-discovering what your team already knows and starts answering with your own decisions, ownership and history.
We think the difference between a wiki and a memory is the difference between what happened and why. Documentation drifts the moment the code moves. A graph that forms from the work itself stays true, because it is linked to the code it describes.
Trust is built into the shape of the product. The engine ships to you and runs on your own infrastructure, writing to a database you control. Your source code and your graph never reach us. We hold only the metadata needed to rotate a leaked key. You bring your own AI and database keys, and they stay on your side. We ship the engine, not the vault.
Oynix is for engineering teams whose knowledge lives in more heads than it should: teams onboarding new people, inheriting old repositories, or handing their agents a codebase without the context behind it. Local runs free on one machine and makes a single engineer faster. The shared server gives a whole team one memory to read from.
The name is the promise. Oy, Greek for the, and Nix, Latin for nothing. Nothing is lost.