San Francisco-based startup Niobia is betting big on a critical but often overlooked part of battery and chip making: helping engineers understand how and why things go wrong. Its AI software flags production defects and lets engineers ask a chatbot what caused them, aiming to deliver answers in minutes rather than days.
How it works (broadly): The company’s customers place AI agents on their production lines to look for defects, then loop in human engineers to work out what might cause a product to fail. The agents run on self-contained GPUs mounted directly on manufacturing equipment, much like a sensor. They can spot pinhole-sized flaws such as scratches. Niobia’s AI also gathers all the data in one place so engineers can ask a chatbot what went wrong and dig deeper into the cause.
Speed is so critical because chips and batteries are some of the world’s most important products, while also increasingly in short supply. As manufacturers continue to develop and improve their technologies, things inevitably go awry during research and development. Engineers then spend days or longer figuring out a fix before they can restart production. Each flaw or mistake can prove costly — sometimes running into the millions of dollars — and set back manufacturing timelines.
Niobia co-founders Gaurav Jha and Harveen Kaur both have PhDs in chemistry and have worked at chipmakers and battery manufacturers, giving them firsthand experience with the challenges within the industry. In a call ahead of TechCrunch Disrupt 2026, where the company is a Battlefield finalist, Jha explained that while what Niobia does is specialized, it serves a multibillion-dollar market that includes some of the biggest battery and semiconductor makers today.
“It’s extremely niche; maybe people will never know about it,” said Jha, speaking of the battery and chip manufacturing market that his company aims to serve. “But it’s like behind the scenes, this is powering everything… we just don’t see it in the regular world.”
“Every product around you, like glasses or shoes, slippers, wallets, and headphones — these processes touch every single part of it,” added Jha.
Niobia’s agents run on open source models that the company trains further with battery and chip research and other data. That training is what lets them identify pinhole-sized defects. Jha said that Niobia chose open source models largely because its customers don’t want their highly sensitive company information going into a public AI cloud where it might be used to train the model.
“None of our customers want their proprietary data to go to the cloud, even if we tell them it’s safe,” said Jha. The startup also outlines how it approaches cybersecurity, given the high-stakes race among adversarial nations to innovate in the battery and chip space. Niobia, for instance, lets its customers hold their own encryption keys so that the startup cannot decrypt the customers’ data.
Jha said that the company’s priority for now is to focus on three specialties, such as film coating, electroplating (which is coating a surface with metal using an electric current), and fluidized bed reaction (a chemical process where solid particles are suspended in a gas or liquid).
He said that the company also wants its technology to become more autonomous, with the eventual aim of building out a fully self-driving factory for its customers.
If you’re curious to learn more, Niobia will present onstage at TechCrunch Disrupt this afternoon in San Francisco. Join us to catch the company and the 19 startups against which it’s competing as part of our storied Startup Battlefield competition, happening today, Tuesday, through Thursday, October 15.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Zack Whittaker is the security editor at TechCrunch. He also authors the weekly cybersecurity newsletter, this week in security.
He can be reached via encrypted message at zackwhittaker.1337 on Signal. You can also contact him by email, or to verify outreach, at [email protected].