Tapestry, an Alphabet “moonshot” project, aims to help countries around the world navigate dramatic load growth. By optimizing grid management and pricing real-time operational visibility, Tapestry functions like a “Google Maps for electrons,” as the project team describes it.
Tapestry’s partnership with PJM, America’s largest regional transmission organization, is helping bring new generation, including renewables, online through applications like HyperQ, an agentic AI tool that reads through interconnection request applications to confirm that they meet all related requirements and flags any errors. These kinds of tasks normally take weeks to complete; but HyperQ analyzes the applications — some as long as 6,000 pages — in a matter of minutes.
The collaboration comes at a crucial moment for PJM, which serves 65 million people in the mid-Atlantic and Midwest. The RTO faces a huge surge in load growth driven, in part, by data centers, and an enormous backlog of interconnection requests. As of last spring, PJM had 200 GW of renewable power waiting in the queue.
“Reviewing that the applications meet these requirements is not the highest and best use of a power systems engineer’s time. It’s a great place for AI to go to work,” said Page Crahan, general manager at Tapestry, at a recent Where the Internet Lives virtual event.
Those interconnection applications aren’t just long; they’re nuanced. The documents must be on company letterhead, for instance, and often require real, “wet” signatures instead of e-signatures. They also include highly technical schematics.
“It’s incredibly complicated,” Crahan said. “A PhD in power systems is going to…go through sometimes hundreds of pages of material from a project developer that’s demonstrating they own the land, showing their project plan, explaining what they’re going to build, confirming that they have financing.”
AI can process all of that in a fraction of the time.
Beyond HyperQ — which rolls out this year — Tapestry is also bringing its unified modeling approach to PJM. Unified modeling allows Tapestry and PJM to have precise visibility into every facet of the grid and the factors impacting it. It’s an effort that requires pulling in trillions of data points — everything from weather to the condition of assets like transformers to the rate of electron transmission itself. Crahan argues that more data will improve the model further.
“It’s getting better and better every quarter,” she said.
Tapestry’s data-driven innovations extend far beyond PJM. During the live event, Crahan identified two case studies in South America where Tapestry has helped countries smooth the process of adding renewable generation to the grid.
In Chile, the government is winding down coal use as part of its plan to convert 70% of its total energy consumption to renewables by 2030. As more wind and solar sources come online, Tapestry’s Grid Planning Tool allows planners to create hourly models that extend twenty years into the future to help developers plan battery installations and curtail supply when it gets too high.
“We’re helping the operators and the developer make more informed decisions about what to do when certain nodes of the grid are producing more [energy] than we need at that time,” Crahan said.
And in Rio de Janeiro, Mayor Eduardo Paes is establishing an “AI City” in the Olympic Park region. The local government, in partnership with Light SA, Axia Energia, and Elea Data Centers, is working with Tapestry to ensure the city can manage rising power demands from over two dozen new clean-energy data centers once the project is complete.
“If we’re going to invest, as a species, two percent of global GDP in energy infrastructure — that’s more than we invested in the railroad, that’s more than we invested in the space mission — we have got to pull together and do it intelligently,” Crahan said.
This is partner content, brought to you by Google. It borrows from a special live episode of Where the Internet Lives, a podcast produced in partnership by Latitude Studios and Google.
Where the Internet Lives is an award-winning podcast about the unseen world of data centers. Follow on Apple, Spotify, or wherever you get your podcasts.
