Engagement models
We do not publish day rates, because a number without a scope is meaningless. What we can tell you up front is how each engagement is structured and which one usually fits.
A feasibility review before anyone commits to a build. You get a written assessment and a baseline number.
Best when you are not yet sure the project is viable.
A defined deliverable with agreed acceptance criteria. Priced once the discovery work has removed the unknowns.
Best when the requirements are clear and stable.
A named team working as an extension of yours, with overlapping hours and direct access rather than a ticket queue. Billed as a monthly retainer, or hourly where the volume varies month to month.
Best for ongoing product work or evolving requirements. This is what most of our long-term clients use, including teams of up to 30 on a single account.
Annotation and labelling priced by volume, or hourly for exploratory and ad-hoc work.
Best for dataset preparation with a defined output format.
How we price
We will not quote a build from a brief alone. The discovery sprint exists so both sides are pricing the same thing.
Code, models, weights, and datasets are yours. No licensing back to us, no lock-in on inference.
If we scoped it wrong, that is our problem. Change requests are priced separately and agreed before work starts.
Describe the problem and we'll tell you which model fits and roughly what it costs - before you commit to anything.