GridCARE: Solving AI’s Biggest Bottleneck with Intelligent Power Solutions
- David Wright
- 41 minutes ago
- 4 min read
How GridCARE Solves the Urgent Power Crisis
The GridCARE platform addresses the most urgent bottleneck in the current artificial intelligence race: the immediate access to power. By utilizing generative AI and establishing close collaborations with utilities, GridCARE unlocks gigawatts of capacity on the existing electric grid. This innovative approach accelerates interconnection timelines from years to months, enabling the critical infrastructure buildout required to support long-term innovation and sustainable growth for the entire sector.
The GridCARE Perspective on the Current Industry Landscape
The global race to power the boom in AI and cloud computing has collided with an electrical grid that was not originally built for today’s required speed or scale. Data centers are currently facing multi-year delays for interconnection studies and are often forced to wait for costly grid upgrades. Simultaneously, utilities are under immense pressure to deliver new capacity faster without compromising reliability or raising electricity rates for consumers.
As the demand for compute power accelerates, U.S. data centers are projected to consume up to 12% of national electricity by 2028. Yet, interconnection delays of five to ten years have become common, creating a significant stifle on innovation. This bottleneck exists even as the grid itself operates at less than 40% utilization, leaving vast amounts of capacity stranded and unused.
While several companies are addressing parts of the interconnection challenge, GridCARE stands out by bringing utilities and data centers together through a unified approach. Leveraging advanced modeling, GridCARE identifies hidden capacity on existing grid assets and works with both sides to operationalize that capacity with flexibility solutions. In doing so, it enables data centers to connect far faster—without waiting for new infrastructure—and allows utilities to accommodate growth without imposing undue cost or delay. This shared model ensures every unlock is technically sound, commercially actionable, and mutually beneficial, creating the fastest, cheapest path to bring new data centers online.
The Origins and Founding Story of GridCARE
GridCARE originated from groundbreaking research regarding AI and grid physics conducted at Stanford University. Recognizing the urgency of power access for AI data centers, the founding team spun the technology out in 2024 to build a commercial platform, becoming the first company to spin out of Stanford’s Sustainability Accelerator. Within the first year, GridCARE established its core team, validated its approach with a major U.S. utility, and launched its commercial solution.
The company’s founders bring decades of collective experience spanning academia, national laboratories, government, and industry. They have led global technology companies, advanced grid modernization programs, and shaped energy innovation policy at the national level.
The leadership team includes Amit Narayan, Co-founder and CEO, a serial entrepreneur who most recently founded and led AutoGrid, the world’s leading virtual power plant platform acquired by Schneider Electric. He is joined by Dr. Liang Min, Co-Founder and Chief Technologist, who serves as the Managing Director of Stanford’s Bits & Watts Initiative, and Dr. Ram Rajagopal, Co-Founder and Chief Scientist, the Director of the Stanford Sustainable Systems Lab. Additionally, Dr. Arun Majumdar, Co-Founder and Advisor, brings his experience as the founding dean of the Stanford Doerr School of Sustainability and former Director of ARPA-E. Their combination of scientific depth, entrepreneurial execution, and policy insight allows GridCARE to operate effectively across both the technical and institutional dimensions of the power system.
The GridCARE Solution and Technological Vision
GridCARE bridges the gap between utilities and data centers, helping both sides unlock and operationalize the capacity that already exists on the grid. This redefines both the technical foundation of grid planning and the business model for data center-utility collaboration. GridCARE’s platform combines generative AI and grid physics to identify, validate, and activate near-term capacity on existing infrastructure. It complements utility planning tools by revealing where and when additional load can be safely interconnected, while helping data centers understand what’s possible and how to structure their projects to utilize the identified capacity.
For utilities, GridCARE acts as a trusted partner to help accelerate interconnections without compromising reliability, regulatory compliance, or affordability. For data centers, GridCARE creates a faster, clearer, and more collaborative path to power, reducing uncertainty and eliminating years of costly delays. By turning system data into actionable capacity, GridCARE enables utilities and customers to move in sync, building trust and accelerating projects that might otherwise stall for years. Together, these capabilities deliver the fastest and most capital-efficient route to powering AI growth while keeping the grid affordable, flexible, and reliable.
How GridCARE Transformed Operations for Portland General Electric
The impact of GridCARE is visible in Hillsboro, Oregon, one of the fastest-growing data center hubs in the U.S. due to its robust fiber connectivity and temperate climate. With rapid growth in the area, data centers were putting extreme strain on Portland General Electric’s (PGE) grid. The existing grid lacked the necessary capacity, and data centers were being told to wait five to seven years for the utility to make upgrades before they could interconnect.
Enter GridCARE. By working closely with PGE’s planning team, using generative AI and grid physics to identify spare capacity on the existing grid, and optimizing flexibility solutions to unlock the excess capacity, GridCARE enabled 80+ MW to be energized by 2026 and 400+ MW by 2029.
Data centers in the Hillsboro region, which previously thought they would not be able to connect until 2030 or later, are now able to come online as early as 2026. This groundbreaking achievement is the first transmission-level flexible interconnection to be implemented in the world. The benefits are clear: data centers now have access to the fastest and most competitive solution for power, addressing the acute capacity challenge. Where traditional planning practices left significant grid potential untapped, this partnership underscores the GridCARE conviction that advancing data center infrastructure requires not only a technical breakthrough, but also a new approach to ecosystem collaboration.
