TechGlobal.news

General Catalyst backs Series A for Texas startup tackling power procurement for AI data centers

Sofia Reyes ·

General Catalyst led a Series A financing on Monday for a Texas based startup that builds software to plan and procure power for AI data centers, according to people familiar with the deal. The company, which operates in ERCOT and PJM, is raising to expand grid modeling, contract tooling, and developer partnerships as interconnection queues and power scarcity slow AI buildouts.

The timing speaks to an investor pivot that has been building for months. Generalist funds are moving from application layer AI bets to infrastructure that mitigates power bottlenecks, a constraint now dictating deployment timelines and capex efficiency for GPU clusters.

The startup sells to data center developers, cloud providers, and AI workloads that control their own sites. Its platform models nodal prices, congestion risk, and curtailment, then structures hedges, short and long term PPAs, and demand response strategies that target predictable all in cost per kilowatt hour. Contracts are priced as a software subscription with an upside sharing component tied to realized avoided costs, which appeals to customers that want operating expense aligned fees.

In Texas, the product ingests ERCOT market data, generator outage schedules, and interconnection queue updates to forecast basis and scarcity pricing. In PJM, it helps navigate queue reform and capacity accreditation changes that affect the value of firm power. Early customers are said to include independent power producers and two AI campus developers that need visibility on multi year power budgets before committing to land and substation work.

General Catalyst has been active around critical infrastructure, backing defense and healthcare platforms that require reliable compute. Its support here signals that AI capacity constraints are investable through software layers that compress power procurement cycles, not only through capital intensive generation projects.

Lowercarbon Capital and several strategic angels with power trading backgrounds participated, people said. The company plans to use proceeds to hire more quantitative engineers and expand data partnerships with transmission planners, which could strengthen moats through proprietary datasets and model accuracy.

Why it matters for startups is straightforward. As GPUs become easier to source, the binding constraint is shifting to megawatts and interconnects. That dynamic is pulling venture dollars into grid software, structured energy contracts, and site selection tools that can shave months off build schedules and lock in unit economics before concrete is poured.