Mirror Particle is building a 'world model' of human behavior

Startups that promise to predict how humans will behave are having a moment. Over the past year, Simile raised $200 million at a $2 billion valuation, Aaru raised $88 million at a $1 billion valuation, and Humans& announced a $480 million seed round in January at a $4.48 billion valuation before launching Persimmon to model human behavior. Two-year-old, San Francisco-based Mirror Particle thinks the dominant approach behind most of these efforts — prompting or fine-tuning large language models to role-play as a target demographic — is fundamentally broken.

“It’s like bringing a super soaker to Niagara Falls,” says Abhivyakti Ahuja, co-founder and CEO of Mirror Particle, which provides brands with an AI engine that predicts consumer behavior and the reasoning behind it. She argues that LLMs, trained overwhelmingly on written language, model the world differently from how humans experience it — through visual perception, spatial reasoning, and social intelligence — so insights drawn from them are based on what humans don’t notice.

Instead, Mirror Particle is building a foundation model from scratch — a “world model,” in Ahuja’s words — that simulates why humans behave the way they do and how that behavior changes over time. Rather than capture a static snapshot of a person, Ahuja said, the model is meant to capture the changing person: longitudinal data on how people shift, what triggers that change, and to what degree. The model draws on a proprietary mix of client customer data, current events, pop culture, and social media to track a demographic segment as it evolves, with an emphasis on “revealed behavior” — what people actually do — over self-reported survey answers.

Mirror Particle has raised an angel round and says it is close to closing its first venture round. It is also among the companies competing in Startup Battlefield 200, TechCrunch’s startup competition at Disrupt 2026 in San Francisco on October 13-15. Like its rivals, the startup’s early go-to-market focus is on market research and brand strategy, where insight budgets already exist. In one pilot, a well-known pet food brand asked which packaging imagery — chicken, beef, or vegetables — would boost sales; Mirror Particle’s technology found the imagery wasn’t the issue. The brand’s recognizability made it look mass-market and cheap, and sales would plateau until it addressed that perception problem.

Ahuja’s interest in modeling the human brain stems from her background in neuroscience and computer science; she later worked at Amazon Robotics, where she met co-founders Will Song, who built sales personalization engines, and Thomson Yen, who focused on how AI agents understand human behavior. The company’s long-term goal, Ahuja said, is to become a “general layer for anticipating human behavior,” moving from population-level analysis toward individual-level insight.

Why it matters

Mirror Particle is staking its approach — a world model built from scratch, rather than an LLM prompted to role-play — against well-funded rivals like Simile, Aaru, and Humans&, betting that modeling how people change over time will prove more useful to brands than static demographic simulations.