San Francisco-based startup Mirror Particle is developing a foundation model designed to simulate human behavior and its evolution over time. Unlike traditional large language models that rely on roleplaying, the company builds a world model from scratch to analyze motivations, social intelligence, and spatial reasoning.
The startup utilizes a proprietary data set that incorporates client customer data, current events, pop culture, and social media. The technology focuses on revealed behavior rather than self-reported survey data, aiming to provide brands with insights into consumer decision-making and product strategy.
Founded by Abhivyakti Ahuja, Will Song, and Thomson Yen, the two-year-old company has completed an angel funding round and is currently closing its first venture round. The team previously worked at Amazon Robotics, where they focused on robotics, sales personalization, and deep learning.
Mirror Particle is participating in the Startup Battlefield competition at the TechCrunch Disrupt event. The company's long-term goal is to establish a general layer for anticipating human behavior, transitioning from broad population-level analysis to individual-level insights.
