Adults 21–34 who buy ready-to-drink beverages
Which combination of flavor, packaging, and price would get customers to try our next beverage?
Our AI-driven simulations show how and why customers, employees, or populations respond to change
There’s prediction, and there’s simulation.
Prediction tells one type of future. It uses historical data and patterns to forecast what is likely to happen next.
But predicting the future is different from changing it.
Changing the future means testing what hasn’t happened yet: the new and unpredictable, where a leap of faith has historically decided what becomes a stroke of genius or a public apology.
True simulation lets us change the future.
Simile is the simulation company committed to training the safest, most accurate behavioral models in the world by painting a living portrait of the unique people who shape it.
What could you simulate?
Explore how simulation can inform decisions across your industry.
Emerging-affluent urban professionals
Which rewards and benefits would make our credit card a customer’s first choice?
Weekend leisure travelers
Which room upgrades and amenities would make guests willing to pay more?
Casual franchise fans ages 18–34
How should we introduce a new franchise to appeal to existing fans and new audiences?
Suburban parents considering their first EV
Which combination of price, range, and charging benefits would persuade shoppers to consider our electric SUV?
Patients managing chronic conditions
Which messages and support services would help patients stay on track with their medications?
Longtime multi-line households
Which changes to our plans would give longtime customers a reason to stay?
Adults 21–34 who buy ready-to-drink beverages
Which combination of flavor, packaging, and price would get customers to try our next beverage?
Emerging-affluent urban professionals
Which rewards and benefits would make our credit card a customer’s first choice?
Weekend leisure travelers
Which room upgrades and amenities would make guests willing to pay more?
Casual franchise fans ages 18–34
How should we introduce a new franchise to appeal to existing fans and new audiences?
Suburban parents considering their first EV
Which combination of price, range, and charging benefits would persuade shoppers to consider our electric SUV?
Patients managing chronic conditions
Which messages and support services would help patients stay on track with their medications?
Longtime multi-line households
Which changes to our plans would give longtime customers a reason to stay?
Validation comes first
How the population is built
Every Simile population starts with real people. Our proprietary algorithms design large-scale studies daily, combining the resulting data with proprietary human behavior datasets to train AI models that know human behavior better than anyone on earth.
How it is validated
We validate against real humans weekly: over 7,000 evaluations across subpopulations and real enterprise use cases, covering distributional distance, effect size comparisons, and accuracy against observed behavior. Our validations train an AI confidence model that tags every result with a predicted accuracy level so you know where to trust it.
How it is refreshed
We continuously monitor for changes in the world and refresh our models where needed. Because different types of data age at different rates, we track freshness at the data level, re-collect data when we detect drift, and incorporate current real-world context into simulations.
You can bring your own data
Loyalty data, balance histories, telemetry: your data, opt-in and customer-governed, trains your custom model and closes the action loop, proving customers behave as predicted.
Simile is the trusted simulation partner for leading enterprises
“With Simile, we don’t need to fail fast in front of customers, we can fail safely in a controlled environment. Teams can test, learn, and refine until they’re confident it’s ready for the real world.”
Srikant NarasimhanVP Enterprise Customer Experience & Insights, CVS Health
“Simile expanded our qualitative research scope by 15X without losing depth. Speaking with their simulated customers helps us quickly identify areas of opportunity for ongoing research.”
Andrew CroccoHead of User Research, Wealthfront
“Simile is a game changer in how we understand customers — accelerating product while helping teams align faster across the organization.”
Fabricio DoreChief Design Officer, Banco Itaú
“We’re partnering with Simile with the goal to accelerate time to market in our product development life cycle. Simile has been a strong partner in our broader focus on understanding our consumers better.”
Bharathi ViswanathanChief Digital & Information Office, Suntory Beverages & Food
“Gallup’s partnership with Simile is built on a shared commitment to methodological rigor, transparency, and technology-driven innovation. Together, we're expanding how deeply we understand people, while staying grounded in what they actually think and feel.”
Joe DalyGlobal Managing Partner
“In a sea of AI tools, Simile is the first to provide a new way for our private equity team to understand consumers with AI.”
Loften DeprezOperating Executive, Garnett Station Partners
Our founders have pioneered the field of AI-based simulation with foundational academic work at Stanford University
Simile was founded by the 
Joon Sung ParkCo-founder & CEO · Creator of generative agentsStanford PhD who pioneered generative agents and population-scale human simulation, including the landmark “Smallville” study and the first simulation grounded in 1,000 real people. Recipient of Stanford CS’s Arthur Samuel Award for best doctoral dissertation.
Michael BernsteinCo-founder & Chief Data OfficerLeading human-centered AI researcher whose work helped define modern social and interactive computing. Stanford HAI Senior Fellow, Sloan Fellow, Tech for Humanity Prize recipient, and creator of the most-cited research in UIST’s history.
Percy LiangCo-founder & Chief ScientistOne of the world’s foremost AI researchers and founding director of Stanford’s Center for Research on Foundation Models. Creator of SQuAD and HELM, and recipient of the Presidential Early Career Award and IJCAI Computers and Thought Award. behind foundation models, generative agent architectures, and rigorous benchmarking practices.
On the Opportunities and Risks of Foundation Models, Stanford CRFM
Research
The report introduced the term “foundation models” and established rigorous evaluation as essential to managing their broad capabilities and risks.
In Simile today
Simile is building a foundation model for human behavior, with evaluation and uncertainty treated as core model capabilities.
Retail Banking Customers
If switching banks took 10 minutes, who would leave and how many seem satisfied today?
Would switch banks
31%
Satisfied, but still switchable
44%
Satisfied and likely to stay
25%
Moderately high confidence
Electric Vehicle ConsiderersCharging access is the leading barrier, followed by upfront price and long-trip confidence.
Charging access
34%
Upfront price
26%
Driving range
19%
Battery and resale value
13%
Medium confidence
Luxury Experience BuyersPersonal recognition and effortless service most often justify the premium.
Personal recognition
32%
Effortless service
27%
Distinctive access
18%
Exceptional quality
15%
Social Simulacra: Creating Populated Prototypes for Social Computing Systems, UIST
Research
Language-model simulations exposed large-scale behavioral paradigms that emerge when social computing systems are populated.
In Simile today
Our platform turns a research question into comparable scenarios and shows how the same population responds as concepts, pricing, or messaging change.
Which product launch will most grow high-value customers?
Ad 1Trail Momentum82%
Ad 2Sun Boost38%
Ad 3Sky Sprint60%
Generative Agents: Interactive Simulacra of Human Behavior, UIST
Research
Published landmark "Smallville" paper, built a simulated town of 25 agents, and debuted multi-agent architecture with memory, retrieval, reflection, and planning.
In Simile today
These architectural ideas help simulations preserve context and produce coherent behavior across a study.
Generative Agent Simulations of 1,000 People
Research
Interview-grounded agents reproduced participant responses at 85% of human self-retest accuracy and reduced racial and ideological bias versus demographic prompts.
In Simile today
Simile's foundation model is trained on real people, grounding population-scale predictions in human data.
Population
1,000
Finetuning LLMs for Human Behavior Prediction in Social Science Experiments, EMNLP
Research
Finetuning on 2.9M responses from 210 experiments improved alignment on unseen studies by 26% versus the base model, outperforming frontier models.
In Simile today
This is the model direction: predict how anyone behaves in any situation, trained and validated on held-out human responses.
Share of responsesResponse option
