TechGlobal.news

Sequoia backs $52 million Series A for Gridfoil to manage power for AI data centers

Sofia Reyes ·

Gridfoil, a New York based startup that builds software to plan and procure power for AI data centers, said Monday it raised a $52 million Series A led by Sequoia Capital with participation from DCVC and Lowercarbon Capital. The company will use the capital to expand its grid modeling and power procurement products in Texas, Virginia, and Ohio, where interconnection queues and tariff volatility have slowed new compute capacity.

The timing signals an investor tilt toward energy software that can unlock near term compute supply, rather than another layer of AI applications. Power constraints have become a gating factor for GPU utilization and revenue recognition, which ties Gridfoil’s product directly to customers’ unit economics.

Gridfoil sells a subscription that prices by megawatt under management, with add ons for day ahead dispatch optimization and on site generation control. Contracts typically span three to five years, which makes gross margin and retention more predictable than usage based AI tooling, according to investors familiar with the company’s model.

Early customers include two colocation developers building campuses in ERCOT and PJM, plus a pilot with a cloud provider evaluating behind the meter generation at a new facility in Central Texas. The software integrates interconnection studies, nodal price forecasts, and demand response programs, then recommends procurement and scheduling that lowers effective cost per kilowatt hour.

Sequoia’s involvement matters because top tier generalist firms have until recently concentrated their AI exposure in models and application software. Backing grid and power software indicates a thesis that compute growth will be gated by electrons and permits more than by model architecture, and that value will accrue to tools that compress time to power.

Competition spans incumbents like Schneider Electric and Siemens for energy management, as well as startups focused on specific layers such as LineVision for transmission monitoring and Crusoe for behind the meter generation. Gridfoil’s bet is that data center developers want a single planning and procurement system that bridges utility processes, merchant power, and on site assets.

For venture markets, the deal highlights a path for AI adjacent startups that tie directly to capacity expansion and cash conversion. As data center buildouts push into secondary markets and co locate with new generation, software that improves power availability and predictability is becoming a core part of the AI infrastructure stack.