This session tells the story of why GPI was launched and built in response to a growing industry gap. FlavorWiki began by helping brands run custom, real-world product experience studies; over time, the repeated lesson was that each project created valuable evidence, but too much of that learning stayed locked inside project files, separated by methodology, market, or team. Now GPI, as a product of FlavorWiki, is being built as an AI-powered product intelligence engine designed to change that. Its model is straightforward: collect real-world product and consumer experience data at scale, enrich that learning through AI-supported structuring and sensory intelligence, and transform disconnected studies into a connected system of record where learning becomes searchable, comparable, reusable, and cumulative rather than disposable.

The GPI engine combines a state-of-the-art sensory intelligence layer with a large longitudinal proprietary product-experience dataset from our owned global consumer community, giving AI models a richer and more contextual signal than isolated, one-off studies. AI and LLM-powered features can synthesize prior studies, identify recurring patterns, generate insights, surface testable R&D hypotheses, and support predictive product learning. So if a reduced-sugar drink wins in a blind screen but weakens in home use, GPI can use AI-supported learning to identify likely drivers, suggest optimization opportunities, and help teams decide whether to adjust sweetness architecture, improve messaging, or generate the next hypothesis to test.

Daniel Protz

Founder & CEO

FlavorWiki - GPI