Pin down the job
We agree exactly what the tool does, on what data, and what “good” looks like. We check the data is really there before building on it.
Build · 02
One real, working AI tool — built and handed over.
Four to eight weeks to a single AI solution that does a real job, set up in your own systems, with a plain guide your team can run it from. Built to a standard that keeps working after we've gone.
What you get
A tool that works — and a team that can run it.
The point of a pilot isn't a demo. It's something real, in your hands, that keeps working once we leave.
- 01
One AI tool that does a specific, agreed job — built and set up in your environment.
- 02
Proper engineering under it: tested, watched for problems, and documented — not a fragile demo.
- 03
A plain operating guide your staff can actually work from, without us.
- 04
A hands-on handover session with the people who'll run it day to day.
- 05
An honest note on what it costs to run, and what it would take to grow it.
- 06
A thirty-day check-in once it's live and in real use.
How it works
Four to eight weeks, with checkpoints you can see.
You see working progress at known points — not a black box that opens at the end.
Build in the open
We build it in short cycles with checkpoints you can try, so it's shaped by real feedback and holds up in real conditions — including where power and signal are unreliable.
Set up & hand over
We deploy it in your environment, walk your team through the guide, and make sure they can run and change it on their own.
Investment
Priced to the build, billed by milestone.
We agree the scope and the price up front, then bill against checkpoints — so you're never paying ahead of the work.
Price$18k–$45kA tight, focused tool sits near the floor; a fuller build near the top.
Duration4–8 weeksFrom kickoff to a working, handed-over tool.
Good to knowOften fundable as a technology or capacity-building line in a grant. Grant-funded builds welcome.
Included
- Scoping and a data check before we build
- The build itself, tested and documented
- Deployment in your environment
- A plain operating guide
- A hands-on handover and a 30-day check-in
Not included
- Ongoing operation or a support retainer (that's the Fractional AI Lead)
- Running AI fully offline / on your own hardware (that's Sovereign AI Deployment)
- New data collection or major data clean-up, if that's needed first
Questions
What people ask before a build.
Do we need a readiness assessment first?
Not always. If you already know the tool you want and the data it needs, we can go straight to a build. If you're not sure yet, the assessment is the cheaper way to get clear first.
What happens after the pilot?
You own a working tool and the guide to run it. Some teams take it from there; others keep us on part-time to grow it or run more builds. There's no lock-in either way.
Will it work where our connection is bad?
We design for that from the start where it matters — low signal, unreliable power, varied devices. If the tool needs to keep working with no internet at all, that's a Sovereign AI Deployment, and we'll flag it early.
Who owns what we build?
You do. The tool, the guide and the know-how to run it are yours at handover.
Related
Before, alongside, and after a build.
Know what you want to build? Let's scope it.
A short call is enough to size the work and give you a real number.