I meet a lot of founders. Most of them are building a wrapper around someone else’s model and calling it a product. Every so often I meet one who is solving the harder problem underneath, and it is worth writing about. Ayush Gupta, the founder of Genloop, is one of those.
He came to a side event I was running during Snowflake. We got talking, and what struck me was how he spoke about his customers. Not as logos on a slide. As people with real problems he clearly cared about solving. You can tell quickly when a founder is in it for the customer rather than the round, and he was.
That care showed up in how he talked about the actual problem. He sees the gap most enterprises are about to fall into more clearly than most. Everyone has rushed to plug a general-purpose model into their business. It works in a demo. Then it meets the messy reality of company data, internal vocabulary, and rules that cannot be broken, and the cracks show. A public model knows the internet. It does not know your business.
Genloop is built around that distinction. The company works on custom and private large language models that are adapted to a specific organization, its data, its language, and its workflows. The idea is not just to host a model safely. It is to make the model behave like it actually belongs inside your company. There is a difference between a system that runs securely and a system that thinks in your terms, and Ayush is one of the few founders who talks about that difference plainly.
The part I find most convincing is where Genloop puts governance. For most teams, governance is the thing they bolt on after something goes wrong. Genloop treats it as part of the foundation. When the model is trained on your data and aligned with your controls, oversight stops being a separate project and becomes how the thing works. For any leader who has to answer to a board about AI, that ordering matters more than another point of accuracy on a benchmark.
There is a simple promise underneath the company that I keep coming back to. Give every team the analyst they have been waiting for. Most people in a business cannot write a query or wait three days for the data team to answer a question. The work Genloop is doing points at a future where they just ask, in plain language, and get an answer they can trust. That is a real shift, and it is closer than most people think.
I am not going to pretend I can predict which AI companies last. Nobody can right now. What I will say is that the founders worth backing are the ones working on the hard, unglamorous layer rather than the shiny surface. Ayush is working on the layer. The conversation I had with him was the kind that stays useful for weeks, where you keep finding new edges to what was said.
If you lead data or AI inside a company, Genloop is worth a look. And if you ever get the chance to talk with Ayush, take it. You can find the company at genloop.ai.
I will keep writing about the founders I meet who are building the parts that matter. He is one of them.
