50/FIFTY

Today's stories, rewritten neutrally

AI6d ago

AI Startup ThinkLabs Raises $28M to Address Power Grid Challenges

ThinkLabs AI secured $28 million in Series A funding to develop AI models that simulate electrical grid behavior amid surging electricity demand.

Synthesized from 5 sources

ThinkLabs AI, a startup developing artificial intelligence models to simulate electrical grid behavior, announced it has raised $28 million in Series A funding led by Energy Impact Partners. Nvidia's venture capital arm NVentures and Edison International also participated in the oversubscribed round.

The San Francisco-based company focuses on using physics-informed AI to model power grid operations in real time, compressing engineering studies that traditionally take weeks or months into minutes. CEO Josh Wong, who previously founded and sold grid software company Opus One Solutions to GE in 2022, said the platform can complete month-long grid studies in under three minutes while maintaining over 99.7% accuracy.

The funding comes as U.S. electricity demand is projected to grow 25% by 2030, driven largely by AI data centers, electric vehicle charging, and building electrification. This surge is straining a grid engineered decades ago for different demand patterns, creating bottlenecks in the complex power flow calculations utilities need to evaluate new connections and infrastructure changes.

ThinkLabs' technology addresses a specific technical challenge: three-phase AC power flow analysis that examines every node on the electrical grid to determine power levels, line flows, and voltages. The company trains its AI models on outputs from established physics-based engineering simulators that utilities already use, then validates results against those same tools to ensure reliability for critical infrastructure decisions.

The company, which spun out from GE Vernova in April 2024, is working with more than 10 utilities and doubled its customer base in the first quarter of 2026. ThinkLabs demonstrated its capabilities through a collaboration with Southern California Edison, showing it could process a full year of power-flow data across more than 100 circuits in under three minutes, work that previously required 30-35 days.

Wong plans to use the funding to advance the product to enterprise grade and expand use cases across utility operations. The backing from Energy Impact Partners, which counts more than half of North America's investor-owned utilities as supporters, provides direct access to potential customers in the conservative utility industry where procurement cycles typically extend for years.

Sources (5)

Bias Scale:
LeftCenterRight
? · Unknown
Financial TimesMar 31, 2026, 9:36 AM
Why do politicians want AI to go faster?
0 · Center
68Trust
Financial TimesMar 31, 2026, 4:01 AM
Why do politicians want AI to go faster?
0 · Center
78Trust

Comments

No comments yet. Be the first!