The startup’s Series B funding round values Simile at $2 billion and will help it develop what it called in the post “a first-of-its-kind confidence model that predicts the accuracy of every simulation.” The company’s goal is to simulate all 8 billion people on Earth.
“With AI, anyone can now create a product, a campaign, a policy or a script,” the post said. “The bottleneck has moved upstream. The hard question is no longer whether you can create something, but rather what to create, for whom, and how to bring it to life. These are fundamentally human decisions that should not be left to chance or handed off to an algorithm to decide.”
CVS Health turned to Simile last fall, The New York Times reported Thursday. Millions of patients never fill or don’t finish their prescriptions, and the health solutions company wanted to figure out how to get people to take their medicine.
The company sought answers from data collected from 400,000 people; only these people were “agentic twins” of actual people created using AI, with Simile handling the data, the report said.
Simile developed its initial model using two-hour interviews with 1,000 people chosen to represent the larger population. The company designed its core model using that data, testing responses against new answers from the human panelists, according to the report.
Meanwhile, financial institutions are replacing live customers with AI-generated stand-ins, or synthetic profiles that can test new products, such as credit cards.
“The synthetic consumer doesn’t just compress timelines,” PYMNTS reported June 22. “It changes how banks bring products to market.”
U.S. Bank deploys synthetic audiences to model consumer segments, testing messaging and refining campaigns before launch. JPMorganChase generates synthetic financial data to simulate market behaviors for risk management and product design. NatWest, Monzo and Santander are using synthetic data ecosystems to train AI models.
However, synthetic data is not inherently safe.
“It can leak sensitive signals through inference and linkage risks,” the report said. “It can also replicate and scale historical biases, embedding them behind a layer of abstraction that makes them harder to detect, audit and challenge.”
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.
