Mirror Particle builds AI model to predict consumer behavior

Mirror Particle is building a foundation model that predicts consumer behavior and explains the motivations behind it, according to TechCrunch. The San Francisco based startup focuses on how people change over time rather than using language models to represent fixed demographic profiles.

The model combines customer data with current events, pop culture and social media. It emphasizes revealed behavior, meaning what people do rather than what they report in surveys, and tracks how experiences alter motivations and constraints.

Mirror Particle initially offers the system for market research and brand and product strategy. Its prediction engine can assess whether a demographic wants a product and identify factors affecting demand beyond advertising or packaging choices.

The two year old company has raised an angel round and is close to completing its first venture round. Its founders are Abhivyakti Ahuja, Will Song and Thomson Yen.