Clear scope, fixed fees, and an exit after every step.
You should know what you are buying, what it costs and how you will judge it before you sign anything. Here is how we make that true.
Three steps. Stop after any of them.
- 30 min · free
Readiness call
We ask how you operate today, where the friction is, and what you have tried. You get a straight view on whether AI helps yet, and which service fits if it does.
You get: a short written follow-up with our recommendation.
- 1–2 weeks
Discovery & proposal
For builds, we look at your systems, data and volumes. For GPU work, we size the workload. Then we send a fixed-fee proposal with scope, timeline, success measures and exclusions.
You get: a proposal you can compare like for like.
- 2–14 weeks
Delivery in phases
Work is split into phases with a review at the end of each. If a phase does not meet its measures, you can stop and keep everything delivered so far.
You get: working outcomes, documentation and a clean handover.
How we price.
Fixed fee per phase
Advisory, governance and build work. The price is agreed before work starts and does not grow unless the scope does, in writing.
Usage-based running costs
Agents and hosted models are charged per minute, per conversation or per GPU-hour, modelled during scoping from your real volumes.
Optional retainer
Monthly tuning, monitoring and support for agents, models and GPU clusters, cancellable with notice.
What we need from you.
- A sponsor with authority to decide
- Access to the systems involved, ideally test accounts
- Real examples: calls, documents, tickets, datasets
- A few hours of your team's time each week
Start with a conversation, not a proposal.
Thirty minutes with someone who has built this before. If we're not the right fit, we'll say so and point you to who is.