Stellenbosch runs two economies that rarely share a developer: the estates and the university. One needs software that survives a harvest, the other needs a research prototype turned into something that can carry paying users. Both fail for the same reason: they are built by whoever was available, and nobody stays to operate the result.
Wine is not an abstract domain here
We build and run Sommelio, our own wine app: photograph a bottle, get a verdict matched to a taste profile that trains itself from every confirmed match. Label recognition from a phone camera, tasting data that is messy by nature, recommendation logic that has to stay useful after a thousand entries. Those are problems already solved once, at our own cost, in production. That is a different starting point from a discovery workshop about wine.
From prototype to product
The spin-out case is the mirror image. The science is sound, the notebook works, and the architecture was never intended to carry a customer. An architecture review is usually the cheaper first step: it establishes what has to be rebuilt before the next round, and what can be left alone. From there it is either a rebuild to a fixed price or a handover plan your own team can execute.
The deal
- First Version (MVP) from $27,800 in two to four weeks
- App Check $2,000 when a prototype or codebase already exists, credited in full if you continue within 60 days
- Fixed price and scope in writing before the first line of code
- Offline-capable where the work happens in the vineyard, not at a desk
- Full handover of code, cloud accounts and documentation
Nearby: app development in Cape Town.