Stanford spinout Simile raises $200M to scale AI behavioral simulation
The Palo Alto startup reaches a $2 billion valuation as it builds synthetic populations to replace traditional market research.
Simile, a Palo Alto-based AI startup, has raised $200 million in Series B funding to advance its platform for simulating human behavior. The round values the Stanford spinout at $2 billion post-money, signaling intense investor interest in the application of generative agents for enterprise decision-making.
The Series B was led by Greenoaks, with participation from Index Ventures, Hanabi, A*, Bain Capital Ventures, CVS Health Ventures, and Definition. This capital injection follows a $100 million Series A led by Index Ventures in February 2026, bringing the company's total funding to at least $300 million in just five months.
The science of synthetic populations
Founded by Joon Sung Park, Michael Bernstein, Percy Liang, and Lainie Yallen, Simile's technology is rooted in academic research. The company's foundation models are based on Park's 2023 "Generative Agents" study—often referred to as the "Smallville" research—which demonstrated that AI agents could use memory and planning to coordinate complex social activities.
Unlike generic AI personas, Simile develops "agentic twins" and synthetic populations designed to mirror real-world human behavior. By grounding these agents in actual data, the company allows enterprises to create a digital mirror of their target demographics to test how real people might react to new products or policy changes.
Redefining market research
Simile is positioning behavioral simulation as a critical new layer of the enterprise AI stack. The goal is to move beyond static surveys and expensive live experiments by allowing companies to run thousands of simulations against synthetic populations before a real-world deployment. This approach aims to significantly reduce the financial risk and time associated with high-stakes business decisions.
By replacing or augmenting traditional market research, Simile seeks to provide a more dynamic and scalable way to predict consumer behavior. The ambition is vast; CEO and co-founder Joon Sung Park stated that the company's mission is to "simulate all eight billion people on earth, accurately."
The road ahead
As Simile scales its operations with this new capital, the industry will be watching whether synthetic populations can truly replicate the unpredictability of human psychology at scale. While the funding suggests strong confidence from venture capital, the primary challenge remains the accuracy of these simulations when applied to complex, real-world economic environments. The company's ability to prove that "agentic twins" can reliably predict human outcomes will determine if behavioral simulation becomes a standard corporate tool or remains a high-tech curiosity.