Data Center Water Startup Omen AI Pays $16 Million for Rival Aquatic Labs as Liquid Cooling Becomes a Chokepoint

Omen AI, a startup that monitors the water flowing through AI data center cooling systems, is buying its closest competitor. The company said it has acquired Aquatic Labs in a $16 million cash-and-stock deal, uniting the two young firms' complementary sensor technologies under one roof.

The deal responds to an unglamorous but growing problem. Data centers in the United States use an estimated 17 billion to 19 billion gallons of water a year to cool their chips, according to Water Foundation figures, and that number is projected to triple within five years as AI buildouts accelerate. For investors pouring billions into AI infrastructure, water is both a cost line and a political liability.

Today, checking the health of cooling fluid is a slow, manual business: samples go out to a lab, and results come back days later. Omen AI, founded in 2024, builds sensors that read the water in real time — detecting the elements in the fluid, spotting failed valves, and catching bacterial growth that can clog a loop. Aquatic Labs attacks the same problem from a different angle: its solid-state sensors measure pH, alkalinity, conductivity and temperature without consumable reagents.

The two technologies fit together like halves of a diagnostic. Omen's founder Zach Laberge and Aquatic Labs CEO Allan Adams were introduced by an investor focused on filtration, and the companies agreed that merging them into a single sensor would produce what Adams called a far richer picture of liquid conditions. Adams, a physicist who spent years building advanced ocean sensors as an MIT professor and Woods Hole Oceanographic Institution researcher, becomes chief science officer of the combined company; the integrated sensor is slated to ship early next year.

The founders are an odd-couple pairing. Laberge dropped out of high school to start his first company, which ran fluid analysis for the hydraulic systems used in heavy construction machinery. Adams took the academic route into nearly the same problem, just submerged in seawater rather than coolant.

The commercial logic hinges on uptime. When a liquid cooling system goes wrong, operators typically drain the entire loop rather than risk destroying millions of dollars of chips — which means downtime until the system is refilled and stabilized. Continuous monitoring lets operators stretch the life of the same coolant, minimize flushes, and catch valve failures before they cascade. Omen's first sensors will monitor an estimated 250 to 300 megawatts of chips by the end of this year, a slice of the market the company values at $5 billion to $7 billion, and it targets the equivalent of a gigawatt of monitored chips by the end of 2027 — figures provided by the company and not independently verified.

Omen raised a $31 million Series A in June, and its focus on operators who have already financed their data centers is deliberate: the company argues its business tracks committed projects rather than the mood of the AI infrastructure market. As liquid cooling shifts from exotic to standard for high-density GPU racks, the monitoring layer around it is quietly becoming an industry of its own.