San Francisco startup Mirror Particle builds a 'world model' to predict human behavior
Two-year-old Mirror Particle says it is taking a fundamentally different approach than rivals relying on large language models to predict and explain consumer behavior for brands. The company is close to closing its first venture round and will compete at TechCrunch Disrupt 2026.
Over the past year, several startups promising to predict human behavior have attracted major funding, including Simile, Aaru, and Humans&. Joining this space is San Francisco-based Mirror Particle, a two-year-old company whose co-founder and CEO, Abhivyakti Ahuja, argues that the industry's dominant approach — fine-tuning large language models to role-play as target demographics — is fundamentally flawed.
Ahuja contends that LLMs model written language rather than how humans actually experience the world, which relies on visual perception, spatial reasoning, and social intelligence. Fine-tuning such models with relatively small amounts of data, she says, can't meaningfully change how they perceive reality.
A different approach
Instead, Mirror Particle is building a foundation model from scratch designed to simulate why people behave the way they do and how that behavior evolves over time. Rather than capturing a static snapshot of a person, the company aims to track longitudinal change — what triggers shifts in behavior and to what degree, treating a lack of change as a meaningful signal too.
The model draws on a mix of client customer data, current events, pop culture, and social media, focusing on "revealed behavior" — what people actually do rather than what they report in surveys. Mirror Particle's initial customers are in market research and brand strategy, where budgets for such insights already exist.
In one pilot, a well-known pet food brand asked which imagery would boost sales on its packaging. Mirror Particle's analysis found the imagery wasn't the issue — the brand's strong recognition made it seem like a cheap, mass-market product, which was capping sales growth regardless of packaging choices.
Ahuja studied neuroscience and computer science at the University of Toronto, where she was influenced by AI pioneer Geoffrey Hinton. She later worked at Amazon Robotics, where she met co-founders Will Song and Thomson Yen.
Mirror Particle has raised an angel round and says it is close to closing its first venture funding round. The company will also compete in Startup Battlefield 200 at TechCrunch Disrupt 2026, held in San Francisco from October 13 to 15.
