See something working before you commit to the full build.
Large builds are usually approved on a slide deck and a shared hope. A month of real work replaces both with evidence.
A 30-Day Build takes the single riskiest assumption behind your product or AI idea and tests it with something that actually runs, used by the people who would live with it.
Deliverables
- Choose one high-value workflow or product slice
- Define success criteria
- Design the experience
- Build a working pilot
- Test with real users or internal teams
- Deliver the next-stage roadmap
- Ideas that need evidence before a budget is approved
- AI use cases where feasibility is the open question
- Teams who want to see delivery quality before a long engagement
Not sure this is the right starting point? Tell us the situation and we will say so — including if the answer is that you do not need us yet.
What actually happens
Pick one thing
One workflow or one product slice — the one where being wrong would be most expensive. Narrowing this properly is most of what makes thirty days feasible.
Agree what success means
A number or a behaviour we can observe at the end, written down before we build. Without it, everyone judges the pilot against a different private standard.
Design the experience
Enough design to make the pilot genuinely usable. Users cannot give you real signal on something they have to be talked through.
Build it
A working pilot on real data where possible. Not a clickable mockup — the point is to find out what breaks when it meets reality.
Put it in front of people
Real users or the internal team who would actually live with it, doing the actual task rather than a scripted demo path.
Decide with evidence
What the pilot showed, what it cost to run, what would change at full scale, and a roadmap for the next stage — including whether there should be one.
From a pilot to a product
If the pilot earns its place, the same senior team carries it into production — with the architecture decisions made deliberately rather than inherited from a prototype.
See Product Engineering30-Day Build FAQ
Tell us the situation in a few lines. We will come back with whether this is the right starting point, and what it would cover.