These 20- and 22-year-olds raised $5M from YC, General Catalyst to study online behavior using vision AI

Spread the love

Chatter Very little sleep is going on, but at 6am, he believes that he believes he is apologizing for the re -opening, and still refrains from the recent fear of family members and an electric scooter.

Although in a few minutes, the 20-year-old Stanford Snaps Snap Focus, how his co-founder sold a startup at the age of 19, landed on Wi-Combinator and collected 5 million for their next company, Human behaviorThe

Launched just a few months ago, the human behavior is betting that the Vision that has been fought with analysis tools like mixpanel and posthogs can do: How people use their products, why they use their products, including conversion or churning.

Instead of manually rely on tagged events or click stream data, human behavior claims that its AI real user sessions show replays and create insights, answers to the most stressful questions of the product parties without the Instructing Code.

The four -month -old WAC startup has closed the $ 5 million seed round in just two days (which is becoming an ideal for current WAC companies), including General Catalist, Paul Graham, Versal Ventures and Wi -Combinators.

The CEO said, “We could do the financial engineering game because we got more offers with higher evaluation, but we didn’t want it,” said the CEO.

Human behavior
LR: Amy Chaturvedi (CEO), Chirag Kawadia (COO), Skyler JI (CTO)Figure Credit:Human behavior

Chaturvedi met his co-founders, Skyler g And ChirpBoth 22, at a hacker house, he organized an excuse to be with friends after his new year at Stanford in 2023.

TechCrunch event

San Francisco
|
October 27-29, 2025

Their first startup, flour was an e-trade accounting equipment that they bootstrap. Like Chaturvedi, Geo got off from college (left Berkeley) and Kawadia graduated.

Although the WAC was initially suspicious about the potential of the flour market, they said that the team was admitted this year to the Accelerator’s Spring Batch, said Chaturvedi. They did it almost immediately after talking to each customer and interrogating any of the problems they had encountered.

The response was consistent: Although the flour may show that any products are being sold or not, customers wanted to know why. Not just accounting reports, but the necessary analyzes driven by behavioral data.

In this new direction, the team has sold the flour for six images to employer.com, The same company that bought the benchAnd people always went to behavior.

Kaudia explains that companies that use traditional dehydration analysis are often needed to set up event trackers for each button and engineering during engineering, burning, sometimes clicking for several weeks.

For a fast dynamic startup, it is far from the norm. “Even once you get this data, you are stuck with a bigger question of how you users actually interact with your product so you can make it better,” he said.

Session Replays are not new, but until recently computer vision models were not accurate enough to pause on their scales. Now they are doing this to summarize and divide the footage of thousands of hours of human behavior. “When we only watch the video, why do you spend hours to write the code to track the clicks?” G added.

Today, human behavior customers-the most fast running series A and B Startups-some features were used, which bugs were present and which users churned to receive short emails by highlighting what. Since launching four months ago, Chaturvedi has said that the company has been increasing 20% ​​of the month-over month.

Founder re -replace an “unnecessary goldmin” by calling sessions. At this point, human behavior helps teams understand users and squash bugs. Over time, the same datasate can strengthen automatic QA and embeds IT support. Their ambition is to make human behavior a session replay’s datadog, spinning a few dozen products from the same original data.

How founders believe that new technology from ground -up believes that they will accept more established players like mixpanel and posthog. “For some of these companies, it can be difficult to replicate what we have because they cannot support the shift without starting their architecture,” quadrilateral commented.

Leave a Reply

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