Multinational B2C corporation: AI Product Discovery: months of specs turned into two days
For one of Europe's largest consumer corporations, we used AI and vibecoding to cut requirements gathering for a new internal system from months to two days.
The challenge
Our client is one of Europe's largest consumer corporations. It operates across more than six markets, pairs a large retail network with digital services, employs thousands of people, serves tens of millions of customers, and reported over 8 billion EUR in revenue in 2024. At that size, every decision about internal software touches a lot of departments and a lot of people.
They came to us with a clear job: write the brief for a new internal system and decide whether to buy something off the shelf or build it from scratch. They had run this play before, and it was always the same loop. Gathering requirements across departments took months. Waiting for developers to build a first prototype took weeks or months more. Then came the feedback, then more waiting for changes. Round and round. In big companies the bottleneck isn't the development. It's the brief - agreeing on what the software should actually do.
Our solution
We cut that whole loop down to two days. First we pulled the high-level requirements from the client. Then we used AI and vibecoding to build a working prototype. No wireframes, no waiting for ten people to find a slot in their calendars.
We took that prototype to two workshops at the client's office. People from every affected department were in the room, and instead of an abstract document to comment on, they had a clickable app in front of them. They pointed at it and told us what they needed to see and how. We rewrote the prototype live, on the spot:
- "We need this figure for the last three years, not just one." Prompt. Done. - "We don't just want the number of checks, we want their dates too, so we know how often they happen." Prompt. Done.
It's a plain Discovery phase, just stripped of the wireframes and the months of prep. We work straight away with something people can actually touch.
Results
Months of work became two days. The client saved time and money and had a decision-ready brief far sooner: buy off the shelf, or build custom.
The second effect mattered just as much. When people from each department see the prototype right away and get to reshape it themselves, they start treating it as theirs. Engagement went up. And the more people shape the brief, the better the final software gets adopted. AI genuinely changed the game here. For the better, if you ask us.
