About
Where biology and production ML actually meet.
Most computational biology projects don’t stall for lack of a good model, or for lack of biological insight — they stall at the handoff between the two: a statistically sound model that ignores a constraint the biology makes obvious, or a biologically careful analysis that never leaves a notebook.
Phronexis exists to sit in that handoff. Every project under Work starts on a public dataset, for a specific reason: the method has to be checkable by someone who wasn’t in the room, before it’s asked to carry client work.
How engagements run
Scoped work you end up owning.
An engagement starts with a short scoping conversation about the decision the analysis actually needs to support, not a generic capabilities pitch. From there it runs as a sprint, a fixed-scope build, or an ongoing retainer — see Services for what each looks like in practice.
Whatever the shape, the output is something that runs on your own infrastructure once we’re done — a documented pipeline, a trained model with its training code — not a locked platform you keep paying to access.