In 2023, Joon Sung Park led the creation of Smallville, a virtual town populated by generative agents that could remember, plan and interact. The research showed that language models could do more than generate content. They could begin to model human behavior.

Joon developed that work alongside Percy Liang and Michael Bernstein, two of the leading minds in AI and human-computer interaction. Together, they laid much of the foundation for what Simile is now building. Joon brings the vision and force of will. Percy brings deep expertise in models and machine learning. Michael has a rare ability to turn research into products people can use. Lainie brought commercial horsepower. After helping scale companies including Hebbia, Valence and TriplePlay, she joined Simile with the urgency, customer focus and operating ability to turn a technical breakthrough into an enterprise business.

We backed Simile at the Series A because we believed this was a one-of-a-kind founding team. We are doubling down because that early promise is becoming a real company.

Left to right: Alysaa Co, Joon Sung Park, Lainie Yallen, Aaref Hilaly

From research to product

Most synthetic-user products start with a demographic profile and ask a general-purpose model to play the part. Simile starts with real people.

Each “simile” maps one-to-one to an individual and is grounded in structured interviews, past choices and behavioral signals. Those agents can then be assembled into populations, allowing companies to test a product, message, customer journey or strategic decision before taking it to market.

Teams get more than a prediction. They can see the range of likely responses, the reasoning behind them and the confidence in the result. Customer understanding becomes always-on—not a study that begins and ends.

The market is responding. In the five months since launch, Simile has grown revenue fivefold, run tens of millions of simulations for Fortune 100 enterprises and expanded to more than 50 people. At CVS Health, the technology supports a program grounded in 2.9 million consented responses from more than 400,000 participants, helping teams explore customer journeys and business decisions in hours rather than weeks.

Putting humanity back into every decision

At its core, Simile is building an AI simulation platform for human decision-making.

Nearly every consequential decision rests on an assumption about what someone else will do: whether a customer buys, a patient follows a treatment, a student learns, a voter reacts or a community adapts. Today, many of those assumptions are made with partial information and a fair amount of instinct. Simile gives us a way to test them first.

The goal is not to replace human judgment. It is to give judgment a richer, more honest view of the people affected by the decision. That could transform decision-making across business, politics, education and society.

Put simply, Simile makes AI more human.

Today, Simile is announcing more than $200 million in funding at a $2 billion valuation. We are proud to continue backing Joon, Percy, Michael, Lainie and the entire Simile team as they build The Simulation Company.

In 2023, Joon Sung Park led the creation of Smallville, a virtual town populated by generative agents that could remember, plan and interact. The research showed that language models could do more than generate content—they could begin to model human behavior.

Joon developed that work alongside Percy Liang and Michael Bernstein, two of the leading minds in AI and human-computer interaction. Together, they laid much of the foundation for what Simile is now building. Joon brings the vision and force of will. Percy brings deep expertise in models and machine learning. Michael has a rare ability to turn research into products people can use. Lainie brought commercial horsepower. After helping scale companies including Hebbia, Valence and TriplePlay, she joined Simile with the urgency, customer focus and operating ability to turn a technical breakthrough into an enterprise business.

We backed Simile at the Series A because we believed this was a one-of-a-kind founding team. We are doubling down because that early promise is becoming a real company.