Emerald AI raises $150M Series A at $1.05B to make data centers grid-flexible

The round closed August 25, 2026, and was oversubscribed. It follows a $24.5M seed led by Radical Ventures and a $25M strategic round in March 2026, after the company launched in July 2025. Total funding is now above $220M, and 12 Fortune Global 500 companies are investors. The core product, Emerald Conductor, schedules AI workloads against the batteries and generation available onsite. The money goes toward scaling commercial deployments worldwide. Over the past year the company ran five demonstrations on live commercial load in Arizona, Illinois, Virginia, Oregon and London. It has since deployed at a full data center in California. Silicon Valley Power, the utility serving Santa Clara, has set up a Flexible Load Interconnection Program with Emerald. Emerald estimates flexible operation could unlock more than 100 gigawatts on the existing U.S. system. Established demand response vendors such as Enel X and AutoGrid serve industrial users, while Emerald targets AI data centers specifically. For AI companies, power access is increasingly what limits how fast capacity comes online. A utility that can curtail part of a load during peak periods may approve interconnection sooner and at larger scale. Teams planning large compute commitments should ask providers whether they offer flexible-load arrangements, and what that means for workload priority and performance guarantees.