Amogh Chaturvedi, a 20-year-old Stanford dropout, is leading a fast-growing startup called Human Behavior, which has raised $5 million in seed funding from top investors including Y Combinator, General Catalyst, Paul Graham, Vercel Ventures, and others. The company, launched just months ago, is using vision AI to analyze online user behavior in ways traditional analytics tools like Mixpanel and PostHog have struggled to achieve. Chaturvedi and his co-founders—Skyler Ji, 22, and Chirag Kawediya, also 22—met in 2023 at a hacker house Chaturvedi organized to live and build with friends after leaving Stanford. Their first venture, Dough, was an e-commerce accounting tool they bootstrapped. Though Y Combinator initially doubted its market potential, the team was accepted into the spring batch after promising to pivot. They did so quickly, after gathering feedback from customers who wanted to understand not just what was selling, but why. The insight led them to shift focus toward behavioral analytics. After selling Dough for six figures to Employer.com, they launched Human Behavior with a bold mission: use computer vision to analyze real user session replays and extract meaningful insights without requiring engineers to manually code event tracking. Traditional analytics tools rely on predefined events and tags, which can take weeks to set up and often fail to capture the full picture of user behavior. Human Behavior’s AI watches session replays, identifies key actions, detects bugs, and flags churn signals—automatically summarizing thousands of hours of user footage into digestible reports. The founders believe the timing is right. Advances in vision AI now allow for accurate, scalable analysis of screen recordings—something that wasn’t feasible just a few years ago. Their customers, primarily fast-growing Series A and B startups, receive daily summary emails that highlight feature usage, critical bugs, and user drop-off points. Since launching four months ago, the company has grown 20% month-over-month. The team sees session replays as an “untapped goldmine” and aims to evolve Human Behavior into a platform that powers automated QA, embedded support, and other products built on behavioral data. They view their approach as a competitive advantage. Unlike legacy analytics platforms, which are built on older architectures, Human Behavior is designed from the ground up around AI-driven insights. Chaturvedi argues that for established players, replicating this model would require a complete overhaul. Despite receiving multiple offers with higher valuations, the founders chose to raise at a lower valuation to align with investors who shared their long-term vision. Their goal is to become the Datadog of session replay—turning raw user behavior into a foundation for a suite of intelligent product tools.