Data centers running AI workloads are facing an unexpected enemy: bacteria growing inside the cooling systems that keep GPUs from overheating. To maximize computing power, operators push chips hotter by changing the liquid coolant to contain more water, which absorbs heat better. But that same water creates ideal conditions for bacterial growth, leading to system clogs that force five to six hour shutdowns costing millions of dollars per incident.

Why Is Bacterial Growth Such a Big Problem in AI Data Centers?

The cooling systems in modern data centers use a specialized liquid mixture of water and bacteria-inhibiting chemicals to keep densely packed GPUs running. When operators increase the water content to improve heat absorption, they inadvertently create a breeding ground for contamination. The bacteria clog the cooling lines, forcing technicians to flush the entire system and take racks offline for extended periods.

Today, most data center managers have no real-time visibility into the chemical health of their cooling systems. They rely on manual sampling, extracting fluid and sending it to laboratories for analysis, a process that leaves them flying blind to emerging problems. By the time contamination is detected, it has often already caused significant damage.

How Can Data Centers Monitor Cooling System Health in Real Time?

Omen AI, a startup founded by 24-year-old entrepreneur Zach Laberge, has developed a tiny spectrometer device that monitors cooling fluid chemistry continuously, spotting bacterial growth before it becomes catastrophic. The device can also detect wear in pumps by identifying copper or chromium particles, and seal degradation by spotting silicon.

The company announced a $31 million Series A funding round led by Nava Ventures, with participation from venture firms CRV and Hard Launch Capital, as well as backing from Vanderbilt University and industrial suppliers Mann+Hummel and Starhill Holdings. Omen has already raised $40 million total since its founding in 2024 and is working with a dozen data center customers to refine its offering.

• Real-Time Detection: The spectrometer continuously monitors fluid chemistry rather than relying on manual lab samples, providing immediate alerts when bacterial growth begins.
• Predictive Maintenance: By detecting metal particles and seal degradation early, the device helps prevent unexpected equipment failures and unplanned downtime.
• Cost Avoidance: Preventing a single five to six hour shutdown can save data center operators millions of dollars in lost compute capacity and revenue.

Laberge's path to solving this problem is unconventional. He founded his first company at age 14 in 2020, raising $3 million to install sensors on construction equipment, and dropped out of high school to pursue the venture. After that startup shut down, he founded Omen in 2024 with a focus on fluid system monitoring for heavy machinery.

The transition to data centers came naturally. Caterpillar, a major supplier of gas-powered turbines and generators for on-premises data center power, is a key customer through its dealership network. About six months ago, those dealerships began asking if Omen could help monitor cooling systems on the data center building side.

"You're not risking huge amounts of downtime because you have no insight into what's going on chemically," explained Zach Laberge, CEO and founder of Omen AI.

Zach Laberge, CEO and founder, Omen AI

Who Is Already Using This Technology?

Omen is working with TensorWave, a company building an AI compute cloud on AMD chips, as one of its early data center customers. TensorWave's president emphasized the critical importance of fluid monitoring in large-scale infrastructure.

"The fluid running through these massive systems is a critical variable that most of the industry is flying blind on. Omen sees the future of infrastructure exactly the way we do, better monitoring to optimally support compute customers," stated Piotr Tomasik, president of TensorWave.

Piotr Tomasik, President, TensorWave

Omen is not alone in recognizing this market opportunity. Pyxis, an established water-monitoring firm, rolled out its own data center coolant monitoring product earlier in June 2026. The emergence of competing solutions suggests that fluid monitoring is becoming a critical infrastructure concern as AI data centers scale.

What Technology Makes Real-Time Monitoring Possible?

Recent advances in optical technologies and signal processing software have made on-premises fluid analytics economically viable at scale. Hardware costs have dropped low enough that deploying sensors across multiple data center locations makes financial sense, while improved signal processing algorithms can extract meaningful insights from noisy sensor data.

"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," said Laberge.

Zach Laberge, CEO and founder, Omen AI

The timing of Omen's funding round reflects broader investor confidence in infrastructure solutions for the AI boom. As hyperscalers like Microsoft, Amazon AWS, Google, Meta, and Oracle continue building out massive data centers packed with GPUs, operational reliability becomes increasingly critical. A single contamination event in a large facility can cascade into significant revenue loss, making preventive monitoring an attractive investment.

Cory Rellas, a partner at Nava Ventures who sits on Omen's board, highlighted the startup's unusual advantage in a traditionally slow-moving industry. "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," Rellas noted.

As AI compute demand continues to surge, data center operators are discovering that managing the physical infrastructure is just as critical as managing the software. Bacterial contamination in cooling systems represents a hidden cost of the AI boom, one that startups like Omen are now positioned to address at scale.