INTERVIEW
summary
Claims to model human taste by combining science-aided AI with fundamentals from chemistry, biology and physics so predictions carry built-in explainability, not just black‑box correlations. Describes building proprietary flavor chemistry datasets and models that represent chemical makeup and flavor structure, packaged as a SaaS for F&B (software product sold to food companies rather than ingredient manufacturers). Positions approach against product-focused firms like Impossible Foods and reports positive early reception at the conference.
AI-written summary · about
related talks
36mAI for human and the planet: AI in nutrition
04mCES Interview
40mFood, Wine and Machine Learning: Teaching a Bot to Taste
03mInterview
10mMy AI Thinks I'm Eating My Feelings (and Other Nutritional Insights)
04mVitaly Gordon Interview
42mQA for ML: How we can trust AI with Food Sustainability
45mContent recommendation at iFood scale