AI startup Vectorial lands $3M in funding to simulate human behavior for consumer research

Vectorial AI Inc. has raised $3 million in seed funding for a platform that tries to do consumer research without recruiting a room full of humans. The startup says its system simulates user behavior so product and marketing teams can get feedback while they’re still building, not weeks after the fact.

The round was led by Smilegate and SenseAI Ventures, with support from the ecosystem funds of the University of California at Berkeley and Stanford University. Angel investors from Meta Platforms Inc., Mastercard Inc. and Shopify Inc. also backed the company. That investor mix makes sense. Vectorial is aiming at a pain point that almost every product team knows too well: research is slow, awkward and often too small to be useful.

Co-founder and CEO Taranveer Singh says most companies still depend on old-school surveys that can take weeks and usually pull in only 50 to 100 people. Smaller startups often skip research entirely and gamble on what they think users want. Vectorial’s answer is a “living model” of target customers that updates as trends and behavior shift, then spits out feedback from thousands of simulated users inside the company’s workflow.

The platform is built around a model called SAPIENS, trained on latent socio-demographic, psychographic and behavioral traits drawn from a database of more than 2.5 million real consumers. Chief Product Officer Rahul Garg said SAPIENS is 30% more accurate than general-purpose models at predicting consumer behavior. Vectorial says it has already used the system with companies in e-commerce, gaming and healthcare, including GoodRx Inc., Boat, Everkind Inc. and Concrete Software Inc., with customers running an average of 600 simulations per month.

There’s still a long way between a demo and a standard. Singh says the company will use the new money to expand its behavioral dataset and build SAPIENS 2.0, while also pushing the idea that AI should sit inside the product development loop, not just outside it as another research tool.