The story behind the stack.
Researcher and developer focused on making local AI accessible on consumer hardware. Runs the TurboQuant research fork, entropy-adaptive KV cache work, and maintains this config generator.
Unicorn assistant. Handles testing, documentation, visual QA, and making sure Lyndon doesn't lose track of context. Also known to have opinions about line-height values.
Because chasing every new AI release is a fast track to losing your mind. We focus on what actually works on real hardware — staying productive, not frantic.