He wrote his first line of code at twelve, shipped three apps before he could vote, and turned down Cornell to wire up the plumbing of modern finance. Now his AI Accountants close the books for some of the biggest companies on earth.
Founder & CEO, Kinter.ai // Co-founder, Alloy Automation (YC W20) // San Francisco
Gregg Mojica builds the connective tissue nobody applauds. The kind of software that, when it works, disappears. Today that work has a name and a mission: Kinter.ai, the company he runs as founder and CEO, sells "AI Accountants" - software agents that execute the month-end close, the unglamorous monthly ritual where finance teams reconcile every number before the books can shut. Enterprises like Amazon, Mastercard, and UPS are named among its customers. The pitch is blunt: let the machine do the close.
Kinter did not appear from nowhere. It grew out of Alloy Automation, the integration platform Mojica co-founded with Sara Du in 2019 and ran through Y Combinator's Winter 2020 batch. Alloy started life as something people described as "Zapier for ecommerce" - a drag-and-drop way for merchants to wire their shops to their tools without writing code. It grew into a unified API and embedded integration layer used to connect hundreds of SaaS applications, then pivoted toward the moment everyone else was racing to: integration infrastructure for AI agents. The same plumbing that once synced an order to a spreadsheet now feeds the agents doing the accounting.
"Our integrations allow you to seamlessly connect your Buy with Prime data with your ecommerce platform."
That is the throughline of his career - take a tedious, breakable, high-stakes seam between two systems and make it reliable enough that a Fortune 500 finance team will trust it with their numbers. It is not a flashy ambition. It is a durable one. And it is the kind of problem that rewards someone who has been doing this, quietly, since middle school.
Look closely and the shift from Alloy to Kinter is less a pivot than a logical next move. Alloy spent years learning how to make hundreds of different applications speak the same language - a unified API, an embedded integration layer, a marketplace of pre-built connectors. That is exactly the capability an autonomous accounting agent needs. An AI Accountant cannot close a month if it cannot reach the ledger, the bank feed, the billing system, and the expense tool, and trust that the data arriving is clean and current. The hard part of agentic finance was never the model. It was the wiring. Mojica spent half a decade building the wiring before the application even existed.
A tech nerd who would rather build the thing that lasts than the thing that trends.- The throughline of Gregg Mojica's work
Most founders romanticize their origin story. Mojica's needs no embroidery. He began coding at twelve. By the time his peers were filling out college applications, he had three apps in the wild - each one a small lesson in shipping, distribution, and what people actually click on.
An anonymous social networking app that pulled 165,000 downloads. A teenager's crash course in virality, moderation, and how fast attention moves.
A screen-sharing tutoring service - real-time help between students and tutors, built before "remote learning" was a phrase everyone knew.
A platform that rewarded students for getting good grades. Equal parts product and behavioral experiment, aimed squarely at his own generation.
The admission letter arrived. The decision did not take long. Instead of a campus, Mojica chose a problem: payments. He went to build at First Data, the payments giant later folded into Fiserv, working as an application architect on API development and the developer portal that outside engineers would build against.
It was an unfashionable move, and a formative one. Payments infrastructure is where reliability is not a feature but the entire product. A dropped transaction is a real dollar. That early apprenticeship in systems that absolutely cannot fail is visible in everything he has built since - and earned him stints across the fintech world at companies including Elavon, with technical writing for AppCoda along the way.
There is a quiet logic to skipping the diploma. The lesson he wanted - how money actually moves between systems, how an API earns the trust of the engineers who depend on it - was not in a lecture hall. It was inside the machine. He went and got it. By the time he started a company, he had already spent years inside the exact category of problem he would go on to solve for everyone else.
Mojica and Sara Du met through the open-source developer community. In 2016, Du invited him to speak at a conference - the first thread of a partnership that would run for years. The two compared notes on what they kept running into: integration and automation were a persistent, expensive headache for commerce companies. Everyone needed their systems to talk; almost no one had a clean way to make it happen.
They built a prototype and launched it on Product Hunt in 2019. The early version did something general-purpose tools could not: it understood ecommerce. It could loop through every line item inside an order and act on each one, the sort of domain-specific muscle a horizontal automation platform never bothers to grow. That focus is what gave a two-person side project room to become a company. Y Combinator's W20 batch followed. When COVID arrived mid-program, they made a characteristically pragmatic call: skip the spectacle of demo day, keep shipping, keep serving the early customers. The team scaled remotely and internationally. In February 2022, Alloy raised a $20M Series A led by a16z's David Ulevitch, pushing total funding past the $25M mark and, eventually, beyond $30M.
Through it all Mojica owned the architecture - multi-region, serverless where it made sense, Kubernetes and EKS under the hood - the unglamorous engineering that lets a small team serve merchants all over the world without the seams showing. He has since turned that hard-won knowledge outward, speaking on app architecture, multi-region scaling, and the grind of SOC 2 certification. He also became an angel scout for Bain Capital Ventures in 2021, writing small checks into the next generation of builders.
Spend time with Mojica's public talks and a pattern emerges. He is fixated on the stuff that decides whether a company actually works at scale: serving a global audience without lag, paying only for the compute you use, surviving a sudden burst of traffic without a single dropped request. He talks about Kubernetes and consumption pricing the way other founders talk about growth hacks - because to him, that is the growth hack.
It is a worldview you can trace straight back to a twelve-year-old debugging his own apps and a teenager choosing payments over a prestige diploma. The instinct is always the same: go to where the problem is least glamorous and most load-bearing, and make it boringly dependable. In the age of AI agents, where software now acts on its own, that instinct is suddenly the whole ballgame. An agent is only as trustworthy as the integrations beneath it.
The partnership with Sara Du tells you the rest. She handled the company as CEO; he held the architecture. Two people who met in the open-source world, kept in touch across his payments years and her stints at DoNotPay, Harvard, and Snap, and then built something together that outlasted the trend cycle - from ecommerce automation, through the unified-API era, into AI. Most startups that old have been acquired, pivoted into oblivion, or quietly wound down. Theirs kept finding the next version of the same problem and going after it. That is not luck. That is two founders who picked a problem deep enough to keep paying out for years.
What does not change is the posture. He keeps choosing the load-bearing wall over the paint. He keeps shipping. And he keeps telling other engineers how he did it - the multi-region tradeoffs, the cost math of serverless, the slog of getting SOC 2 - because the knowledge is only worth something if it travels. For a founder whose product is meant to be invisible, he is unusually willing to show his work.