As AI compute demand pushes data centers to get more performance from every rack of GPUs, one overlooked operational issue is moving into the spotlight: the health of the fluids running through those facilities.
Omen AI is building around that problem. The company says a tiny spectrometer can monitor fluid health in real time, giving data center operators earlier visibility into chemical changes and bacterial growth before contamination turns into downtime.
Why coolant has become a data center risk
Liquid-cooled chips rely on a fluid mixture made of water and a substance that inhibits bacteria growth. The mix matters because water absorbs heat better, which makes it useful when data center managers want to run chips hotter.
That tradeoff can create a new problem. When the fluid contains more water, bacterial contamination becomes more likely. If that contamination builds up, it can clog the flow through the cooling system.
The current response can be expensive and disruptive. Operators may need to flush the system, and that can mean shutting down a rack for five or six hours at a potential cost of millions of dollars.
For AI infrastructure, that is not just a maintenance nuisance. When GPU capacity is valuable, a rack taken offline for hours represents lost compute availability at exactly the point when customers are trying to squeeze more output from the hardware they already have.
What Omen AI says it can see
Omen AI’s core product is a small spectrometer designed to watch fluid chemistry continuously. The company’s pitch is that data center operators should not have to wait for problems to become visible through system performance or lab reports.
CEO and founder Zach Laberge framed the issue as a lack of chemical visibility. "You’re not risking huge amounts of downtime because you have no insight into what’s going on chemically," he said.
The same monitoring concept began outside data centers. Laberge started Omen in 2024 with a focus on fluid systems in construction machinery, where real-time awareness could replace the slower process of extracting samples and sending them to a lab.
The device is not limited to bacterial growth. In heavy equipment, Omen’s system can identify signs of wear by detecting materials in the fluid. Copper or chromium can point to pumps and pumps wearing out, while silicon can point to seals.
How the startup moved toward AI infrastructure
Omen’s early customer base included Caterpillar dealerships for its heavy vehicles business. That connection helped point the company toward data centers, because Cat is also a major supplier of gas-powered turbines and generators used to provide on-premises power for those facilities.
According to Laberge, the shift started when dealerships began asking whether Omen could help beyond turbines. "That was kind of the transition," Laberge told TechCrunch. About six months ago, "a lot of the dealerships were saying, ‘Hey, we’re starting to put sensors on our turbines, can you guys do anything on the building side of things?’"
Omen then found that data center buildings contain fluid systems across more than one area, including HVAC systems and chip cooling. That opened a broader market than the company’s original construction machinery focus.
The company is now working with a dozen data center customers as it builds out its offering. One of them is TensorWave, which is building an AI compute cloud on AMD chips.
Piotr Tomasik, TensorWave’s president, described coolant as an under-monitored part of large systems. "The fluid running through these massive systems is a critical variable that most of the industry is flying blind on," he said in a statement. "Omen … see the future of infrastructure exactly the way we do, better monitoring to optimally support compute customers."
The funding and the market signal
Omen AI said it raised a $31 million Series A round. The round was led by Nava Ventures and included participation from CRV, Vanderbilt U niversity, Mann+Hummel, Starhill Holdings, and Hard Launch Capital, along with personal investments from executives at Bridgestone, GM, Johnson Controls, and TensorWave.
The company has raised $40 million since its founding in 2024. For a startup built around industrial fluid monitoring, the data center shift shows how AI infrastructure demand can pull technologies from slower-moving equipment markets into faster-growing compute environments.
Laberge’s founder path is also unusual. He founded his first company in 2020 when he was 14, raised $3 million to install sensors on construction equipment, and later dropped out of high school. After that startup shut down, he started Omen.
Cory Rellas, a partner at Nava Ventures who sits on Omen’s board, said customer conversations helped validate the company. "It’s rare to see such a young founder who has the respect of established, large corporations in a space that moves a bit more slowly," he said. "For Omen in particular, much of our diligence came through our introductions with large customers which quickly validated their approach."
Why real-time monitoring is arriving now
Many organizations still rely on mailing fluid samples to labs when they need insight into fluid condition. Omen is not the only company trying to bring that analysis on-site. Pyxis, an established water-monitoring firm, rolled out its data center coolant monitoring product earlier this month.
The timing is tied to progress in optical technologies and signal processing software. Laberge said the combination has made the approach more practical: "Hardware is just cheap enough that it makes sense to play at scale, and then signal processing lets us make more sense out of the noise," he said.
For data center operators, the underlying idea is straightforward. If fluid problems can be detected earlier, operators may have more time to respond before bacterial growth, wear, or contamination turns into a shutdown. In an AI compute market where every rack matters, seeing what is happening inside the coolant loop could become part of keeping the broader system online.