Niv-AI has emerged from stealth with $12 million in seed funding to address growing power inefficiencies in AI data centers, where fluctuating GPU demand can force operators to reduce usage by up to 30%. The Tel Aviv-based startup, founded by Tomer Timor and Edward Kizis, is developing sensor-driven systems to measure and manage GPU power consumption at millisecond precision.
The company aims to help data centers better align with grid constraints while maximizing utilization of expensive AI hardware. Investor Lior Handelsman of Grove Ventures highlighted the need for new infrastructure approaches as AI workloads scale.
Niv-AI plans to build predictive AI models that synchronize power loads across facilities, positioning its platform as an intelligence layer between data centers and the electrical grid. Deployment in U.S. data centers is expected within months.
Featured image: Credit: Niv-AI
