Gridcare has raised $64 million in an oversubscribed Series A funding round from a combination of tech and energy investors, the company announced today. Sutter Hill Ventures led the investment, which included participation from John Doerr, National Grid Partners, and Future Energy Ventures, among others.
The news comes less than one year after Gridcare emerged from stealth in late May 2025 with $13.5 million in seed funding.
The company uses artificial intelligence to unlock the grid’s underutilized capacity. It employs grid physics, collecting billions of data points across public and proprietary sources — such as utility planning models, interconnection queues, permits, rates, and extreme weather data — to generate a granular map of the available grid capacity. Gridcare then validates its model and assumptions directly with utilities before using the map to help data center developers find available power for their facilities, reducing interconnection timelines to months rather than the usual years it can take to get connected.
CEO and founder Amit Narayan told Latitude Media last year that ideally the company “enters the process before traditional site selection, fundamentally changing how developers approach location decisions,” from a land-first model to a power-first model.
The company has grown rapidly in the last year, as the wider industry has embraced its thesis that power is the primary driver for data center development. This strategy has found a ready market as the sector is increasingly defined by speed-to-power constraints and by the need to squeeze as much power as possible from existing infrastructure. Accordingly, AI applications for the grid are becoming essential tools for both utilities and developers.
As Alaina Bookstein, SVP and head of business development at Gridcare, noted on stage at Transition-AI last month, running “a quadrillion scenarios” to create such a granular map of the grid is something that “you simply couldn’t do even 18 months ago; there wasn’t the compute power to do it.” Today, it allows both utilities and developers to “have open eyes,” because “you all of a sudden have a data set to have a productive conversation.”
These are the conversations that can encourage willing data centers to act as flexible grid assets, a practice that is slowly but surely gathering steam as the strain of AI boom on the existing energy system becomes more obvious.
“We are offering [data centers] the ability to build business cases in a way that was not possible without some of these AI platforms like Gridcare,” Bookstein said. Thanks to Gridcare data, a data center developer can figure out in advance the amount of flexibility it needs to bring to the system and plan around that number, so “they can think about leveraging VPPs in a way that is not this amorphous ‘We’re going to call you to curtail at any time.’”
The model was validated in October of last year, when Gridcare helped Portland General Electric access a granular viewpoint of its own system and accelerate large load interconnections in Hillsboro, Oregon, a rapidly growing data center market.
“By optimizing what’s already on the grid today, and just creating better visibility and partnering very closely with the utility, we were able to create actual unlock for five different data centers,” Bookstein said. “It is actual, credible new interconnection agreements with flexible commitments for a total of 80 megawatts this year, ramping up to 400 MW in ‘29.”
Gridcare is currently engaged in similar projects to accelerate time to power for AI data centers for a total of over two gigawatts of new AI compute capacity in over a dozen markets, according to today’s announcement.
