Emerald AI raised $150 million in a Series A funding round led by JERA Ventures, the company said in an announcement, valuing the startup at $1.05 billion. JERA, the power generation and trading unit of Japan's Tokyo Electric Power Company, joined multiple institutional and strategic investors in the round.
Emerald AI builds software that lets data centers adjust their power consumption based on real-time grid conditions. The platform allows operators to shift compute workloads to times when electricity is cheaper or more abundant, cutting both operating costs and strain on grid infrastructure. The startup said its technology targets 100 gigawatts of untapped U.S. grid capacity that could support AI infrastructure without requiring new generation or transmission assets.
Data center operators face twin pressures: soaring electricity costs as AI workloads intensify, and grid constraints in regions where new power plants take years to build. Emerald AI's approach lets operators make trade-offs between speed and cost by running flexible workloads during off-peak hours or when renewable generation peaks. The software layer sits between the data center and its utility, reading grid pricing signals and supply forecasts.
JERA operates coal, gas, and renewable generation across Japan and Australia, giving it direct experience managing grid-scale demand shifts. JERA said the investment aligns with its strategy to support technologies that optimize energy use in high-demand sectors.
Emerald AI was founded to address a specific technical problem: most AI workloads run on fixed schedules regardless of power prices or grid conditions. That inflexibility has made data centers a growing burden on regional grids during peak hours, particularly in Texas and California. Grid operators in both states have begun curtailing data center expansion permits or imposing constraints on new connections until transmission capacity expands.
The $150 million Series A follows earlier funding from venture firms and cloud infrastructure investors. The round values the company at 7 times its Series A check size, a multiple typical for infrastructure software at this stage. JERA joins a growing set of energy companies, including Shell and bp, that have deployed venture capital into power-tech startups solving grid and cost problems.
Emerald AI must demonstrate that its software can cut data center grid demand by meaningful volumes while meeting customer performance guarantees. If the platform attracts enough operator adoption to shift even a few gigawatts during peak hours, it could reduce wholesale electricity prices in congested markets. The company will need to prove its model works across multiple regions and data center types before reaching the scale JERA and other investors are betting it can achieve.
