Three steps. No discovery theatre.
Every engagement runs the same short arc: a conversation, a working prototype, then production. You can stop after any step and keep what has been built.
There is no phase that only makes sense if you buy the next one, and no deliverable whose purpose is to justify the invoice for the phase before it.
- Discovery
- Free
- Prototype
- Days
- Pricing
- Phased
What an engagement asks of you
- 3
- Gates you can stop at
- 0
- Discovery invoices
- 20min
- Before anything is built
- 2–6weeks
- To production
Keep what has been built at every one of them.
The call that scopes the work is free, and stays free.
The whole commitment at the start of an engagement.
From a signed prototype to a system in real use.
Where the time actually goes.
The decisions are short and the build is long. That is the right way round, and it is the opposite of how most engagements are shaped.
Three gates. Every one of them is an exit.
You are never buying the whole thing to find out whether the first part works. At each gate the exit is drawn as clearly as the continuation.
- 01After discoveryStop here and keep
A ranked shortlist and an honest read on feasibility
Costs you nothing
orPrototype the top candidate - 02After the prototypeStop here and keep
A working proof against your data, plus its numbers
Fixed price, agreed before we start
orTake it to production - 03After the first releaseStop here and keep
A running system, its documentation and its repository
Quoted from what the prototype revealed
orWiden the scope, or stop here
Discover
A short call to find where AI has the biggest leverage in your business.
We ask about the work, not the technology: what takes the longest, what gets redone, where the errors show up. Most conversations surface two or three candidates. We tell you which one we would start with and, more usefully, which ones we would not.
What we need from you
- Someone who knows how the work actually gets done today
- A rough sense of volume — how often, how long, how many people
- Permission to ask blunt questions about what goes wrong
You walk away with
A ranked shortlist and an honest read on feasibility.
Free. No follow-up sequence, no deck.
Where this can go wrong
The most common outcome of a discovery call is that the idea you arrived with is not the best one available. That is a good result, and it costs you twenty minutes.
Prototype
A working proof of concept, so you decide on evidence rather than promises.
We build the narrow version of the real thing against your actual data. It runs, it can be wrong in ways you can see, and it is measured against cases you recognise. This is where most assumptions — ours included — get corrected cheaply.
What we need from you
- A representative sample of real data — the messy kind, not the clean export
- Twenty or so cases you already know the right answer to
- One person who can review output and say whether it is good
You walk away with
Something you can use, plus numbers on how well it works.
Small, fixed price, agreed before we start.
Where this can go wrong
Sometimes the prototype shows the accuracy is not there yet at an acceptable cost. We report that plainly rather than tuning the demo until it passes.
Ship & scale
Production deployment, integration, and ongoing improvement.
Integration into your systems, access control, monitoring and a rollout your team is prepared for. Then the part most projects skip: watching it in real use, fixing what reality exposes, and widening the scope only once the narrow version has earned it.
What we need from you
- Access to the systems it needs to read from and write to
- A decision on where it runs and who administers it
- A named owner on your side once it is live
You walk away with
A system in production with someone accountable for it.
Quoted once the prototype has told us what it involves.
Where this can go wrong
Real usage always surfaces cases the test set missed. We budget for that period rather than treating it as scope creep, which is why the first release is deliberately narrow.
What the first week actually looks like.
Assuming you can get us a data sample and a decision-maker. Those two things, not the build, are what usually set the clock.
- Day 0
Twenty-minute call. We leave with a shortlist and you leave with an honest read on feasibility.
- Day 1
A one-page scope for the prototype: what it does, what data it needs, what 'working' will mean.
- Day 2–4
Build. You get a link to something running, not a status update.
- Day 5
Review against your own cases. Numbers on the table, and a decision that is yours to make.
Priced in phases, because that's the only honest way.
Nobody can quote a production AI system accurately before seeing how it behaves on your data. Firms that do it anyway are either padding heavily or planning to renegotiate later. We would rather price the part we can see.
Discovery
FreeTwenty minutes and a ranked shortlist. No obligation in either direction.
Prototype
FixedA small, agreed sum for a working proof against your data. Priced before we start, and it does not move.
Production
QuotedScoped from what the prototype revealed, in stages you can approve one at a time.
Gate one is a twenty-minute call.
Free, 20 minutes, and you leave with a ranked shortlist either way.
Start with the twenty minutes.
A free 20-minute call, no strings attached. We will tell you where the leverage is — and where it is not.
- No sales deck, no discovery invoice
- You get a ranked shortlist either way
- If AI is the wrong tool, we say so