Shourya Vir Jain left electrical engineering at IIT Delhi to build RamAIn, a computer-use agent that clicks, types, and reads any screen. His bet: the boring, legacy corners of the enterprise are where automation actually matters.
Most people who spend a summer inside a global consulting firm come away with a slide deck and a stack of frameworks. Shourya Vir Jain came away with an annoyance. He kept watching sharp, expensive analysts do something almost comically low-tech: clicking through the same legacy screens, copying a number from one window, pasting it into another, over and over, because the software running the actual business had no clean way to talk to anything else.
That annoyance became a company. Jain is the co-founder and CEO of RamAIn, a San Francisco startup in Y Combinator's Winter 2026 batch. The product is a computer-use AI agent - software that operates other software by simulating a mouse and keyboard, reading whatever is on the screen and acting on it. The company's line for what it does is refreshingly blunt: “If you can see it, RamAIn can do it.”
It is a small idea with a large surface area. Almost every enterprise runs on tools that predate the API era - procurement portals, insurance claim systems, pharmacy operations software, revenue-cycle tools that were last redesigned when flip phones were current. You cannot bolt a clean integration onto most of them. But you can teach an agent to use them the way a person does.
Jain has a name for the thing that bothered him at McKinsey. He calls it the judgment tax. In a Forbes Technology Council essay published in June 2026, he laid it out plainly.
The pattern is everywhere once you notice it. A traditional automation script handles the clean 80 percent of cases well. The other 20 percent - the weird invoice, the mistyped field, the vendor who formats things differently - falls to a senior analyst who fixes it by hand, every single time. That person's involvement is the tax. And the more you automate the easy part, the more the hard part concentrates on your best people.
His critique of the older approach - robotic process automation, or RPA - is that it never really solved this. “RPA only works when inputs are perfectly clean and follow rigid rules,” he wrote, “which is almost never true in enterprise environments.” The moment a button moves or a field shifts, the script breaks, and someone has to file a ticket to fix it.
RamAIn's answer is an agent designed for messiness rather than against it. Instead of a fixed script that assumes the screen never changes, it watches, decides, and adapts - handling unexpected UI changes on its own, keeping a human in the loop for the calls that matter, and leaving an audit trail of every decision so a compliance team can see exactly what happened.
The deeper argument Jain makes is not about technology at all. It is about ownership. For decades, the people who understood a business process and the people who could automate it were different people, sitting in different departments, speaking different languages.
He pushes leaders to reframe the whole question. “The more productive question for operations leaders today is not which technology to adopt next,” he writes. “It is which process would benefit most from giving ownership back to the people running it.”
To understand why Jain thinks in whole systems, it helps to know how he spends his time when he is not shipping. He is a FIDE-rated chess player at 2118, and was once ranked in the global top 20 for players under 16. He is also a certified karate black belt. Both pursuits reward the same thing: patience, pattern recognition, and the discipline to see several moves ahead before touching a piece.
His own one-line description of himself reads like a resume for a very particular kind of person - “student at IIT Delhi, rated chess professional, certified karate black belt, and passionate about productivity, social issues and entrepreneurship.”
The two met as students and decided not to wait for graduation. They moved to San Francisco, joined Y Combinator, and started raising - RamAIn is backed by Y Combinator, General Catalyst, Nexus Venture Partners and PeakXV. For a company still counting its employees on two hands, that is a notable roster of names.
There is a certain contrarian streak in choosing legacy software as your battlefield. The flashy demos in AI tend to involve brand-new apps, clean data, and greenfield problems. Jain went the other direction, toward the tools that make people sigh - the twenty-year-old desktop client running a hospital's billing, the procurement portal with a layout only muscle memory can navigate.
That is also where the money and the pain concentrate. The industries RamAIn targets - procurement, insurance, healthcare operations, finance, retail, pharmacy, logistics - are exactly the ones drowning in repetitive interface work that no API can reach. A second-time founder who has already felt the sting of building without customers tends to aim at where the pain is loudest.
Whether RamAIn becomes the company that finally retires the copy-paste workflow is an open question, and a hard one. Plenty of well-funded firms are circling the same idea. But Jain's framing has a clarity that is easy to underrate. He is not selling a futuristic vision so much as pointing at a chore everyone recognizes and asking why a human still has to do it. For a founder who learned to see the whole board before making a move, that is a familiar kind of patience: play the position in front of you, not the one you wish you had.