Sobre aquesta sessió
Ask a chatbot to design a perfume and it will hand you one in seconds - over budget, past regulatory limits, and built on materials nobody can supply. That gap between a plausible formula and a viable one is why formulation has resisted automation far longer than most people assume, and it's where the real modelling problems live. We open with a short talk on where AI is genuinely creating value in formulation today across fragrance, flavour and specialty chemicals, and where it still falls over. Then you build: connect your own AI assistant to a simulated fragrance R&D environment - ingredient palette with olfactive and cost data, reference formulations, a scoring engine - and design a perfume against a brief, iterating until it holds up. Two tracks run side by side, so you can work entirely in plain English with the assistant you already use, or ignore our tooling and build your own model against the same data. Who it's for: business and technical alike: anyone curious about how AI is applied in industrial R&D. No chemistry background needed. Senior engineers who want to build their own approach will find a track for it. Requirements: own laptop and access to an AI assistant (Claude, ChatGPT or equivalent). Python and an IDE only if you take the advanced track. Setup instructions sent ahead of the session.