Technical News

These children aged 20 and 22 collected $ 5 million in YC, general catalyst to study online behavior using the AI ​​vision

Amogh Chaturvedi runs out of little sleep, but a lot of conviction at 6 am, he is groggy, an excuse for the reprogrammed and always in shock of a reluctant fear involving a family member and an electric scooter.

In a few minutes, however, Stanford’s stall, 20, approaches how he and his co-founders sold a startup at 19, landed in Cominator and collected $ 5 million for their next business, Human Behavior.

Launched just a few months ago, human behavior bets that vision AI can do what analysis tools like Mixpanel and Posthog fought with: Give companies a real understanding of the way people use their products, including why they convert or trigger.

Instead of relying on manually marked events or click data, human behavior affirms that AI looks at real user session reruns and generates information, answering the most pressing questions of the teams without hours of instrumentation code.

The four -month startup YC has closed its $ 5 million seed round in just two days (which becomes a standard for current YC companies), with donors including General Catalyst, Paul Graham, Vercel Ventures and Y Combinator.

“We could have done the financial engineering game because we obtained more offers with higher assessments, but we did not want this,” said the CEO.

LR: Amigh Chaturvedi (CEO), Chirag Kawedia (COO), Skyler Ji (CTO)Image credits:Human behavior

Chaturvedi met his co-founders, Skyler Ji and Chirag Kawediya, both 22, in a pirate house that he organized in 2023 as an excuse to live and build with friends after his first year in Stanford.

Techcrunch event

San Francisco
|
October 27-29, 2025

Their first startup, Dough, was an electronic commerce accounting tool they exploded. Like Chaturvedi, Ji abandoned the college (leaving Berkeley) while Kawediya graduated.

Although YC was initially skeptical about the potential of the Dough market, the team was admitted to the accelerator spring lot this year, assuming they would end up, says Chaturvedi. They did it almost immediately, after talking to all the customers and informed about any other problem they have faced.

The comments were consistent: although the dough could show which products were selling or not, customers wanted to know why. Responding to this analysis required behavioral data, not just accounting reports.

With this new management, the team sold dough for six figures to use.com, the same company that bought the bench and did everything on human behavior.

Kawediya explains that companies using traditional analyzes often need engineers to configure event trackers for each button and click, hours of burning, sometimes weeks, engineering time.

For a fast startup, it is far from ideal. “Even once you have this data, you are always stuck with the biggest question of how users really interact with your product so that you can improve it,” he said.

Session reruns are not new, but until recently, computer vision models were not precise enough to analyze them on a large scale. Now they are, and human behavior does it to summarize and segment thousands of hours of sequences. “Why spend hours writing code to follow the clicks when we can just watch the video?” Ji adds.

Today, Human Behavior customers – mainly series A and B startups quickly – get daily summary emails highlighting the features used, which bugs have appeared and which users have shot. Since its launch four months ago, Chaturvedi said that the company has increased by 20% per month.

The founders call the session restores an “unexploited gold mine”. Currently, human behavior helps teams understand users and buckets to crush. Over time, the same data set could power automated AQ and integrate computer care. Their ambition is to make human behavior the place of data from the replay of session, by turning dozens of products from the same basic data.

Building with new technologies from zero is the way the founders think they will face more established players like Mixpanel and Posthog. “For some of these companies, it could be difficult to reproduce what we have because their architecture cannot support the quarter of work without starting again,” said Chaturvedi.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button