WASHINGTON, D.C., July 1, 2025 — Emerald AI has officially emerged from stealth with a $24.5 million seed funding round, spearheaded by Radical Ventures, and backed by investors including NVIDIA’s venture arm, AMPLO, CRV, and Neotribe. The Series seed also brings on board notable individual supporters like Google’s Jeff Dean, former U.S. Secretary of State John Kerry, Malcolm Turnbull, Fei-Fei Li, and John Doerr.
🔋 Bridging AI and Energy
The startup’s core innovation is its Conductor platform, a real-time orchestration system that enables AI data centers to modulate compute workloads based on power grid conditions. In live tests in Phoenix in collaboration with Oracle, NVIDIA, EPRI, and Salt River Project, the technology demonstrated a 25% reduction in power use over a three-hour peak demand period, while maintaining AI performance.
🌐 Recasting Data Centers as Assets
Traditionally viewed as energy drains, AI data centers using Emerald AI's system can now act as "virtual power plants", dynamically adjusting usage to support grid stability. By intelligently shifting compute tasks, data centers can access power more quickly—sidestepping long interconnection delays—and alleviate grid stress during peak hours.
🔍 Strategic Timing
-
The U.S. energy grid is under growing strain, with data center energy demand projected to increase by 50–100 GW by 2030.
-
Interconnection queues have stretched 7–10 years or more, posing a serious bottleneck for AI infrastructure expansion.
-
Emerald AI targets this challenge head-on by integrating AI compute and grid management via software.
🧠 Leadership & Vision
Founded by energy expert Dr. Varun Sivaram—a Rhodes Scholar and former U.S. energy diplomat—Emerald AI has assembled a seasoned team, including leaders from Amazon, Intel, and solar energy. The company is also part of NVIDIA’s Inception program.
🚀 Next Steps
Emerald AI plans to scale its platform with larger pilots in Phoenix, followed by national rollouts over the next six months. With the backing of its high-profile investor group, the company aims to establish a new standard for grid-responsive AI infrastructure.