AI strategy
Your board wants AI. In 10 days you know exactly what to build.
Engineers with 25+ years in production measure what your work costs today, rank the cases where a model moves that number and price the first one. On day 10 the memo is on your desk, and the build ships inside 30 days.
Tell us what the board asked for. Within 48 hours you have the scope, the price and the date the memo lands.
A board mandate, pilots that never shipped and data nobody has labelled.
- board asks for AIno target
- pilotsnever shipped
- dataunlabelled
- vendor deckscompared
10 days to the memo, most of them spent measuring
The cases worth building with today's numbers, a build-or-buy call on each and one scoped delivery.
- shortlist2–3 use cases
- baselinemeasured today
- build/buy callper case
- first deliveryscoped
From board mandate to scoped build in 10 days
- 01days 1–3
Find where the money goes
Interviews plus everything countable: ticket volumes, handling times, error rates, rework. You get the numbers your own dashboards never showed.
- 02days 3–5
Rank by what can be checked
A case with a recorded right answer can be evaluated. A case without one turns into opinion the day it ships, so it drops to the bottom however good it sounds in the room.
- 03days 5–8
Build or buy, case by case
For each case: what a product costs, what building costs and what you own afterwards. We build software for a living and still name the cases where buying wins.
- 04days 8–10
Scope the first delivery
One case with an eval set, a baseline, a date and a fixed price. Our engineers can start building it the next morning and ship it inside 30 days.
What the platform does, what our engineers do
Reading your systems for feasibility
Platform
What data exists, where it sits and whether the work can reach it.
Measuring the baseline
Both
Extraction is mechanical. What counts as an error gets agreed with the people doing the job.
Ranking the candidates
Engineer
By whether the result can be checked after it ships.
The build-or-buy call
Engineer
Judgement, with our own interest stated in the same paragraph as the recommendation.
Data readiness
Both
The platform reports what is there. An engineer prices what it takes to make it usable.
The memo
Engineer
A short document with every number shown, so your own engineers can check each one.
You walk out with a memo full of numbers and a first delivery you can start the next day.
Advice from the people who build it
The engineers who write the memo build what it recommends, and they have been shipping production systems for more than 25 years. That makes them the best people to write it. Every estimate in the memo is one we will sign as a fixed price the next day, because we have built the same kind of thing many times over.
Take the first 3 days on their own if you want a smaller step. You get the measured baseline in writing, and the document is yours.
What executives ask us first
You build software. How is your advice not a sales pitch?
It is all in writing. The memo names at least one case where buying beats building and says which product. Where our interest touches a recommendation, the memo says so beside it. The first delivery is priced separately, so the memo stands up in front of any other vendor you show it to. Get a quote and see the outline.
Do we need a data platform first?
Less of one than you were told. Most first cases need a few years of records and a way to read them. A warehouse program is a reliable way to spend a year and a half before learning whether the case works. We find out in 10 days. Send the process and we start there.
Our pilots keep dying after the demo. Why?
Nobody defined what correct looks like before building, so the pilot could not be evaluated and had no road to production. The second step of our work closes that gap, and the first delivery ships with an eval set from day one. Send us the pilot that died.
How is this different from your code assessment?
The assessment reads a codebase and plans a rewrite. AI consulting reads a business process and decides where a model pays for itself. A company with an old system and an AI mandate often needs both, and we run them back to back. Tick both in the form.
How much of our people's time does this take?
About 3 hours each from someone who owns the process and someone who approves budget. No workshops and no leadership offsite. Our engineers do the rest and deliver the memo on day 10. Send the form today and the first interview is booked within 2 days.
Ask us where to start
Write down what the board asked for and which processes eat the most hours. Within 48 hours you get the scope and a fixed price. On day 10 the memo is on your desk, and the first build can start the next morning.
An engineer replies first, usually asking what you can already measure.