I spent years developing beers for the Indian market across AB InBev, SABMiller and United Breweries. New product development is a wall of data long before it’s a beer — which idea to back, how to hit the flavour brief, and whether the pilot survives full scale. This is the honest, 3-part story of how AI and analytics crunched that data for me, and exactly where they stopped being useful.
No hype. The data narrowed the field, sped the trials, and tamed the scale-up — but the judgement that picks a winner, tastes the beer, and signs it off stayed human throughout.
The series
- Which Beer Should We Make Next? How I Let the Data Decide — turning consumer and market data into a ranked shortlist of what to brew.
- Crunching Recipe and Sensory Data — learning from every past batch to hit the brief in fewer trial brews.
- From Pilot to Full Scale — predicting the scale-up shift and converging batch by batch.
Read them in order for the full arc of a development project, or jump to the stage you’re stuck on. Every post in the series is tagged #beer-npd.
It pairs naturally with the From Brewer to AI series — the career story behind the same hands-on work.