The most expensive part of predicting human behavior, it turns out, is not the supercomputer. It is the human. You need to find them, pay them, and wait for them to answer your survey, and then you must hope they are telling you the truth. Aaru, a startup founded in 2024, is betting it can cut that cost to near zero by replacing the panel with a simulation. Its method is to spin up thousands of AI agents, give them personas and access to public data, and ask them what they think.
It is a bet that has already attracted over $50 million in venture capital, including a Series A led by Redpoint Ventures that valued the company at a headline $1 billion [TechCrunch, Dec 2025]. The initial proof point was political. In the 2024 US election, the company says it used 5,000 of its AI agents to predict outcomes, reportedly coming within the margin of error and missing the final tally in key races by less than 400 votes [36Kr] [Semafor, 2024]. For a team of founders who were, until very recently, more likely to be studying for a math final than running a board meeting, it was a striking debut.
The teenage founding wedge
Cameron Fink (CEO) and Ned Koh (President) started Aaru at ages 18 and 19, shortly after deciding that college,a night at Dartmouth for Fink, two weeks at Harvard for Koh,was not for them [Vanity Fair, 2025]. They brought on John Kessler as CTO when he was 15 [TechCrunch, Dec 2025]. The team's youth is the most visible risk factor on the cap table, but it is also the source of their wedge. They are native to a world where AI is the default tool for any complex problem, and they approached the staid, expensive world of market research and strategic forecasting as a systems problem waiting for a computational solution. Their product instructs agents through chain-of-thought prompting, mimicking multi-step human reasoning to improve accuracy [Semafor, Sep 2024]. The promise to clients like Accenture, an early investor and customer, is near-instant, bias-free synthetic research that can model complex scenarios without the lag and cost of traditional focus groups [TechCrunch, Dec 2025].
The synthetic research playbook
Aaru's target is the enterprise insight budget historically spent on consultants, surveys, and panels. The initial use cases are in product testing for wealth management and scenario planning for large consultancies. The unit economics argument is straightforward: if a simulated population of 10,000 agents can be created and queried for a fraction of the cost and time of a traditional study, the savings compound quickly. The company's reported ARR was below $10 million as of late 2025, a figure that underscores both the early stage of commercial deployment and the aggressive growth expectations baked into its valuation [TechCrunch, Dec 2025]. The investor syndicate, which includes Felicis, General Catalyst, and Accenture Ventures, is betting that the team can translate a novel prediction engine into a scalable SaaS business for research firms and corporate strategy teams.
| Founder | Role | Notable Background |
|---|---|---|
| Cameron Fink | CEO | Co-founded at 18 after a brief stint at Dartmouth [Vanity Fair, 2025] |
| Ned Koh | President | Co-founded at 19, left Harvard after two weeks [Vanity Fair, 2025] |
| John Kessler | CTO | Joined at age 15; participated in Z Fellows [SF Standard, Feb 2025] |
Where the simulation could stall
For all the promise of its election call, Aaru faces a predictable set of scaling challenges. The market research industry is a relationship business, built on trust and methodological rigor that boards and C-suites understand. Replacing that with a black box of AI agents, no matter how accurate, requires a profound shift in buyer psychology. Then there is the data question. The accuracy of the simulation is only as good as the data used to train and inform the agents; gaps or biases in that data will propagate through the results. Finally, there is the sheer execution lift of going from a brilliant prototype to a hardened enterprise product with the integrations, security, and support that global firms demand. These are not technical problems but classic go-to-market hurdles that often trip up even seasoned teams.
The competitive landscape includes companies like Fairgen, which also uses synthetic data for research, but Aaru's angle is the active simulation of reasoning agents rather than just data generation. The more formidable incumbent, however, is the traditional market research firm itself,the Nielsen, the Ipsos, the Gartner. Their moat is not technology but decades of institutional credibility. For Aaru to succeed, it must not just be more accurate. It must become more trusted.
Do the unit economics work? Run a back-of-envelope check. If a traditional focus group study for a product launch costs a client $50,000 and takes three weeks, and Aaru can deliver a comparable simulation for $5,000 in three hours, the value proposition is clear. The open question is how many of those $50,000 studies are truly replaceable with a simulation today, and how many still require the intangible, messy feedback that only real humans in a room can provide. Aaru's bet is that the former category is vast and growing. Its next twelve months will be about proving that its agents can do more than predict elections,they must predict purchasing decisions, market shifts, and consumer trends with enough consistency to become a budget line item. The company is not just selling software. It is selling a new kind of confidence. To win, it must beat the confidence that a Fortune 500 strategy chief has in the familiar, human-delivered report from a legacy firm.
Sources
- [TechCrunch, Dec 2025] AI synthetic research startup Aaru raised a Series A at a $1B 'headline' valuation | https://techcrunch.com/2025/12/05/ai-synthetic-research-startup-aaru-raised-a-series-a-at-a-1b-headline-valuation/
- [36Kr] 16-Year-Old CTO Leads Team to Accurately Predict US Election with 5,000 AIs, Margin of Less Than 400 Votes | https://eu.36kr.com/en/p/3596704428556551
- [Semafor, Sep 2024] Article on Aaru's chain-of-thought prompting | https://www.semafor.com/article/09/05/2024/aaru-ai-agents-election-prediction
- [Vanity Fair, 2025] As Teens, They Founded a Billion-Dollar Startup. Now They Want to Shape the Future. | https://www.vanityfair.com/news/story/aaru-founders-ai-simulation
- [SF Standard, Feb 2025] High school and college dropouts are rushing to SF for the AI boom | https://sfstandard.com/2025/02/13/founders-dropouts-ycombinator-aiboom-startups/
- [Crunchbase, 2025] Aaru Seed Round | https://www.crunchbase.com/organization/aaru