Key Takeaways
- $28 million Series A closed on March 31, 2026, led by Energy Impact Partners (EIP), a global energy transition fund backed by more than half of North America’s investor-owned utilities
- Strategic investors include NVentures (NVIDIA’s venture capital arm) and Edison International, parent company of Southern California Edison, signaling deep hardware and utility conviction
- ThinkLabs AI’s platform compresses traditional grid analysis from weeks to under 3 minutes, running over 10 million scenarios in under 10 minutes at greater than 99.7% accuracy
- The company has now raised a cumulative total of $33 million (including its $5 million seed round in May 2024), positioning it to scale across a U.S. grid market where electricity demand is projected to grow 25% by 2030
Quick Recap
New York-based ThinkLabs AI has officially closed a $28 million Series A financing round, the company confirmed via a Globe Newswire press release dated March 31, 2026. The round was led by Energy Impact Partners LP, with strategic co-investment from NVentures and Edison International, alongside returning backers GE Vernova, Powerhouse Ventures, Active Impact Investments, Blackhorn Ventures, Amplify Capital, and a large undisclosed North American investor-owned utility. The raise comes less than two years after ThinkLabs launched out of GE Vernova with a $5 million seed round in May 2024.
Physics-Driven AI Powers Next-Gen Grid Solutions
ThinkLabs AI, founded and led by Josh Wong, who previously served as General Manager of Grid Orchestration at GE Vernova, has developed a physics-informed AI platform designed to automate grid planning and real-time operations for utilities. Unlike conventional machine learning platforms that operate as black boxes, ThinkLabs grounds its models in the actual physics of how electrical grids work, creating what it calls “physics-informed AI digital twins” that replicate every line, bus, and node of a utility network.
The company’s flagship product, ThinkLabs Copilot, is a digital assistant built on these digital twins. It handles a range of complex grid workflows: interconnection automation, real-time state estimation, contingency management, dynamic operating envelopes, and capacity planning. In practical terms, a monthly-long grid study that traditionally required expensive computation over weeks can now be completed in under three minutes, with the system capable of running 10 million scenarios in under 10 minutes while maintaining greater than 99.7% power flow accuracy.
The proceeds from this round will be directed toward expanding customer delivery, broadening the product portfolio of physics-informed AI agents, and scaling deployment with ecosystem partners. The oversubscribed round is also being read as a vote of confidence in the company’s commercial traction: ThinkLabs has reported active pilots with more than 10 utilities to date.
The Grid Is Breaking Under AI’s Weight
The timing of this raise is not incidental. The same AI revolution that is generating excitement across the tech sector is simultaneously creating an unprecedented strain on electrical infrastructure. U.S. electricity demand is projected to grow 25% by 2030, a pace the grid was simply never designed to accommodate, according to consultancy ICF International. Data centers, EV charging networks, and electrified buildings are all adding layers of demand onto infrastructure that was engineered for a different era.
The edge AI in smart grids market reflects this urgency. It was valued at $15.49 billion in 2025 and is on track to reach $19.46 billion in 2026, a compound annual growth rate of 25.7%. Looking further out, the market is projected to hit $48.91 billion by 2030. The broader applied AI in energy and utilities segment is expected to grow from $969 million in 2026 to $5.33 billion by 2035, at a CAGR of 20.84%.
Regulatory pressure is intensifying in parallel. Utilities across North America are being asked to add grid capacity at timelines the industry has never previously attempted. The involvement of NVentures, which brings NVIDIA’s high-performance computing ecosystem into the picture, and Edison International, which manages Southern California Edison, suggests that the largest players in the energy value chain see software-based grid intelligence as a structural solution rather than an incremental upgrade.
Competitive Landscape
ThinkLabs AI’s most direct comparators in the grid AI space are Utilidata and AutoGrid, two companies operating in adjacent but distinct parts of the market.
| Feature / Metric | ThinkLabs AI | Utilidata | AutoGrid |
| Core Approach | Physics-informed AI digital twins for grid planning and operations | Edge AI embedded in smart meters and grid hardware via Nvidia Karman module | SaaS platform for DER orchestration and demand-side flexibility management |
| Total Funding Raised | $33M ($5M seed + $28M Series A) | $126.5M total (Series C: $60.3M, April 2025) | $170M+ across 18 rounds (acquired 2023) |
| Lead Investors | Energy Impact Partners, NVentures (NVIDIA), Edison International | Renown Capital Partners, NVIDIA, Quanta Services | Microsoft Climate Innovation Fund, Shell Ventures, National Grid Partners (pre-acquisition) |
| Key Differentiator | 99.7%+ accuracy; compresses weeks-long studies to under 3 minutes; 10M scenarios in 10 minutes | Hardware-embedded AI at the grid edge; real-time device-level decision making | Large fleet of contracted DERs; multi-utility commercial scale; mature platform |
| Stage | Series A (Early commercialization, 10+ utility pilots) | Series C (Growth stage, active utility deployments) | Acquired by Enel X (mature/exit stage) |
| Focus Layer | Grid planning and operational control rooms | Smart meters and grid-edge hardware | Distributed energy resource management |
| Utility Backing | GE Vernova, Edison International, large North American IOU | NVIDIA, Quanta Services | National Grid, E.ON, Shell, Microsoft |
Strategic read
ThinkLabs AI leads in planning-layer automation, particularly the compression of complex power flow studies and multi-scenario contingency analysis, a workflow no other platform handles at this speed and accuracy level.
Utilidata has a deeper deployment footprint at the grid edge through hardware partnerships, making it the stronger candidate for utilities focused on distributed device management and real-time smart meter intelligence. AutoGrid, now acquired, had the broadest commercial reach but has moved into an enterprise integration track following its acquisition.
TechnoTrenz’s Takeaway
I have been watching the grid intelligence space for a while now, and this raise feels genuinely significant to me for a reason that goes beyond the dollar amount. In my experience, the funding rounds that actually shift an industry are not always the largest ones. They are the ones where the investor mix tells you something the headline number cannot.
Here, you have NVIDIA’s venture arm writing a check into a company that runs on physics-based simulation, not GPU-hungry black-box inference. That is a calculated signal: NVentures is not just making a financial bet, it is planting a flag in the type of AI it wants running the infrastructure its own chips power. I think this is a big deal because the grid problem is not a software problem in the traditional sense. It is an engineering problem that needs to be solved at the speed of software.
ThinkLabs has threaded that needle with physics-informed models that utilities can actually audit and explain to regulators, which is something generic large language model approaches simply cannot deliver in a regulated industry. The 99.7% accuracy figure at grid power flow is not marketing language. It is the threshold utilities require before trusting an automated recommendation that could affect millions of homes.
