This is like giving Symrise’s flavor scientists a super-smart assistant that has tasted millions of recipes. It predicts which ingredient combinations will give the right flavor and work well in a product before anyone mixes them in the lab, so you get to market faster with fewer failed trials.
Reduces time and cost of developing and optimizing new flavor and fragrance formulas, while improving hit rates for consumer-preferred products and cutting down on physical experiments and reformulations.
Proprietary formulation, sensory, and consumer preference data combined with domain-specific predictive models tightly embedded into Symrise’s flavor development workflow.
Hybrid
Vector Search
High (Custom Models/Infra)
Access to high-quality, labeled sensory and consumer preference data; integration with lab systems and ingredient databases; and potential inference latency/cost at very large formulation search spaces.
Early Majority
Focus on AI-driven flavor prediction and formula optimization at scale within a major flavor house, likely using large proprietary datasets of formulations, sensory panels, and consumer tests to guide NPD decisions faster than traditional trial-and-error labs.