Human Behavior uses vision language models to watch web user sessions, analyze behavior, identify key moments, surface issues, and make insights searchable via natural language.
Human Behavior
Our view is that direct AI analysis of visual user session data (as Human Behavior does) will render traditional event tagging and manual replay review obsolete, creating a new category leader. The three co-founders are extremely technical, and we were immediately impressed by their speed.
Current tools like FullStory or PostHog force teams into a frustrating trade-off between understanding what users are doing and why they do it. Human Behavior's AI-native, vision-based approach to session replay analysis is not just an incremental improvement but a fundamental technological shift. Their platform acts as a tireless AI product manager that "watches" every single user session replay. It identifies and labels key behaviors and patterns across all sessions, so product teams can ask complex questions in natural language and receive quantitative answers. They offer a lightweight SDK or can connect to existing replay tools like PostHog, LogRocket, and Hotjar.
Founders
While at UC Berkeley, Skyler built the second largest e-commerce Discord server on the internet. He dropped out from the CS program, to become a CPA and build accounting software (an AI QuickBooks for ecommerce businesses).
Amogh dropped out from Stanford CS. He won math and astronomy olympiads, and co-authored a book on biochemistry and a peer-reviewed paper on collagen synthesis while in high school.
Chirag did Math and CS at the University of Chicago, and dropped out from the Masters program to work with Skyler and Amogh. He co-authored a demonstration paper presented at SIGMOD 2024 (a premier database conference).