The Berkeley engineer who sold a student side-project back to his own university - then went off to help build Uber, and now wants to make every Shopify storefront convert.
Yuxin Zhu spends his days thinking about a deceptively small question: why is it so hard for an online store to put up a page that actually sells something? That question is the whole premise of Replo, the San Francisco company he co-founded in 2021 and runs as CEO. Replo is a low-code platform that lets ecommerce brands design custom, high-converting Shopify landing pages - drag, drop, swap a product, ship - without filing a ticket with an engineer.
It is the kind of problem Zhu has been drawn to his whole career: a boring piece of friction that everyone tolerates until someone refuses to. He has built that thing twice now. The first time, he was a freshman who could not find a decent course.
Replo's pitch is narrow on purpose. It targets mid-market ecommerce companies - brands selling somewhere between $1 million and $100 million a year - the ones big enough to care deeply about conversion rates but not big enough to keep a full design-and-engineering team on staff just to publish marketing pages. For them, a landing page is not decoration. It is the difference between a campaign that pays for itself and one that quietly loses money.
In 2011, Zhu arrived at UC Berkeley to study computer science and met a classmate named Noah Gilmore. By 2012, the two had built Berkeleytime, a search engine for the university's notoriously clunky course catalog. It did one thing well: it made finding and evaluating classes fast.
Students noticed. The site grew to over a million sessions a year and became profitable enough to cover the founders' tuition - a rare thing for a campus side-project. Then came the twist most founders never get: in 2015, the University of California acquired the tool. Zhu's first exit was a sale back to the very institution whose problem he had solved. Berkeleytime still serves more than a million people a year.
The lesson stuck. Find a real, specific annoyance inside a community you understand. Fix it well enough that the people living with the problem start telling each other about it. Worry about the business model second.
Between Berkeleytime and Replo, Zhu put in the reps. He wrote software at Yelp from 2013 to 2015, then joined Uber, where he stayed for six years. He started as a senior software engineer and grew into managing engineering, product and design teams. His teams traveled across four continents launching new Uber products, and he was there through the company's IPO - a front-row seat to what scale looks like, and what it costs.
Six years is a long time to stay at one company in San Francisco. When Zhu finally left, it was not to chase a trend. It was to go back to the thing he was best at: finding a specific, unglamorous problem and building the tool nobody else had bothered to build well.
When Zhu and Gilmore reunited to start Replo, they did not lock themselves in a room and theorize. They took on consulting clients and, in effect, worked for free - hand-building Shopify experiences for design partners. Every awkward request, every workaround, every “can it just do this?” was market research disguised as client work. By the time they shipped a product, they already knew what it needed to do.
It worked. After launching in 2022, Replo onboarded more than 1,000 brands and 300 paying customers in three months - almost entirely on organic word-of-mouth. That growth is also what produced Zhu's most honest public quote: word-of-mouth giveth, and it can taketh away. A company that lives on buzz wants something more durable underneath it, which is why he has talked about building a real community around the product rather than depending on the hype cycle.
In February 2023, Replo announced a $4.2 million seed round. The cap table reads like a who's-who of people who care about design and developer experience: Figma Ventures, Y Combinator, Infinity Ventures, La Famiglia, and Vercel founder Guillermo Rauch, alongside strategic investors from the ecommerce world. The company has since raised over $8 million total, with later backing that includes General Catalyst.
Zhu's ambition for Replo has grown past landing pages. The company now describes itself as building every company's go-to-market specialist - the default way businesses, from ecommerce to SaaS, design and ship web experiences that convert. The through-line from a freshman's course-scheduler is unmistakable: remove the friction between an idea and the thing customers actually see.
His first startup exit was to his own university. The University of California bought Berkeleytime in 2015.
That student side-project was profitable enough to pay his tuition - and still serves over a million people a year.
He met co-founder Noah Gilmore in 2011, both computer-science students at Berkeley. They are still building together.
His Uber teams launched products across four continents before he watched the company go public.
He started Replo by working for free, treating consulting clients as a live focus group for the product.