Simile raises $200M at $2B valuation to build foundational models of human behavior for Fortune 500 retailers and banks

Simile raises $200M at $2B valuation

Simile has raised $200 million at a $2 billion valuation. The company, founded by Joon Sung Park, builds foundational models of human behavior — AI systems designed to simulate and predict how consumers think, choose, and act across market segments.

The pitch is that most AI labs are racing to solve objective problems: coding, mathematics, natural sciences. Simile is pointed at the other half of the human brain — values, preferences, taste, and the distribution of behavior across populations. Rather than training on expert knowledge, Simile partners with market research firms to build representative samples of human populations, then trains models to replicate and predict behavioral distributions at scale. Park says Simile currently represents tens of millions of people in simulation, with a stated ambition of reaching one billion.

CVS is a named customer, alongside Fortune 10 and Fortune 50 retailers, major financial institutions, and CPG companies. The deployment model starts with replication: customers run known studies through simulation, verify the output against ground truth, then extend into pre-deployment testing for new markets, new products, and message testing.

“We raised $200,000,000 at 2,000,000,000. The last time I was at TBPN, I believe it was about five months ago. Since then, the company, literally the team itself, has quadrupled. Simile is a company, it's an applied AI lab that is creating foundational model of human behavior, where we create models that can predict human behaviors across different market segments in the future and the multi agent simulations.”

Product interface

Customers access the platform through a filtering interface that lets them define a target population by any attribute — demographics, geography, behavior — and then query that population directly. Inputs can be text questions, images, video ads, Figma mocks, or live websites. Simile's agents traverse those assets and return feedback as the specified population would give it.

Park argues this isn't just a tool for large enterprises. He frames simulation as a way to democratize consumer insight — giving early-stage companies access to population-scale feedback that would otherwise require a research budget only large incumbents can afford.

Agentic direction

As agentic development pipelines drive the cost of production toward zero, Park sees a second use case opening up. When it becomes trivial to generate 10,000 product variations, the scarce resource becomes knowing which variant people actually want. Simile is building MCP and API integrations so its behavioral models can inform not just human decision-makers but agentic systems acting on delegated authority from humans.

The team has grown 4x in the five months since Park's last public appearance. At $2 billion and 5 months of that growth trajectory, the company is moving fast enough that the $200M round looks less like a milestone and more like fuel.