Week of Oct 9, 2026 Batch Other Tech

Large-load interconnection OS

Startup Idea Brief Published Week of Oct 9, 2026

An audit layer for utility interconnection queues that separates real data center demand from speculative filings before capacity gets allocated.

Large-load interconnection OS turns a utility or grid operator's interconnection queue into an audited, prioritized forecast instead of a pile of developer filings nobody can evaluate fast enough. It ingests queue data, site control records, power purchase and financing history, and developer track records, then scores each request on how likely it is to actually get built, flagging speculative placeholders, duplicate filings across sites, and options developers are holding just to keep a spot in line.

The first customer is the utility or state regulator that now has to defend its interconnection decisions in public, not the data center developer. Texas regulators are running exactly this exercise today, an audit of hundreds of queued data center projects against a hard deadline, with no software built for the job. The product replaces outside consultants and spreadsheets with a system of record that produces the same defensible, line-by-line justification every time a request is approved, denied, or re-ranked.

Once embedded in that audit workflow, the same scoring engine becomes the tool developers and financiers use to judge whether their own queue position is real, turning a compliance product into the de facto pricing layer for interconnection risk.

Why Now

Speculative load has overwhelmed grid planning capacity faster than any previous demand cycle: ERCOT's large-load queue grew from roughly 63 GW at the end of 2024 to an estimated 474 GW by August 2026, and regulators can no longer tell real demand from developers parking requests just to hold a place in line. Texas's pause on new interconnection approvals and its state-mandated audit, plus PJM utilities tying large-load forecasts to actual commitments rather than raw requests, mean grid operators now need a standing, auditable methodology rather than one-off consultant studies. That shift from manual review to systematized, repeatable scoring is a software problem that effectively did not exist at this scale two years ago.

Demand Signal

ERCOT's large-load interconnection queue reached roughly 474 GW by August 2026, with officials citing about 90 percent of that as data center requests, which prompted a pause on new interconnection approvals for projects of 75 MW or larger while the state audits hundreds of filings. Texas regulators followed with a rule requiring large-load developers to post 50,000 dollars per MW of requested capacity before interconnection studies even start, a direct response to queue-stuffing. On the PJM side, utilities have collectively forecast about 55 GW of new large load by 2030, roughly double the new generation capacity currently being planned to serve it, so the same filtering problem is spreading well beyond Texas.

Defensibility

The moat is built from things that compound slowly and cannot be bought off the shelf: live integrations into each grid operator's queue data and each utility's internal systems, a scoring methodology that regulators have actually accepted and cited, and the relationships inside utility planning and regulatory affairs teams that decide which vendor gets trusted with audit-grade work. Every audit cycle a utility or commission runs through the product adds case history that sharpens the scoring model, making it harder for a new entrant to match accuracy without years of the same data. Once a state commission references the tool's output in a public order or rulemaking, switching costs rise sharply because the methodology itself becomes part of the regulatory record.

Market Signals

ERCOT queue mostly speculative: ERCOT's large-load interconnection queue sat around 474 GW in August 2026, with about 90 percent of that cited as data center requests, more than five times the grid's record peak demand. Utility Dive

Texas froze new interconnection approvals: Texas paused approval of new data center interconnections for projects of 75 MW or larger while the state audits hundreds of pending filings. (Utility Dive)

Utilities now charge to even ask: Texas regulators adopted a rule requiring large-load developers (75 MW or more) to post 50,000 dollars per MW of requested capacity before interconnection studies begin, aimed directly at speculative queue-stuffing. White & Case

PJM utilities are out ahead of generation: Wood Mackenzie found PJM utilities have collectively committed to serve about 55 GW of new large load by 2030, roughly twice the new generation capacity currently being planned to support it. Wood Mackenzie

Visibility tools exist, audit tools do not: Point products like DC Hub already track interconnection queue status and data center share across grid operators, but they report raw queue data rather than auditing which requests are real, leaving the verification layer utilities and regulators now need unbuilt. DC Hub

How I'd Build This

MVP Scope

The smallest viable version is a single-grid-operator audit report: ingest ERCOT's public large-load interconnection queue data, cross-reference site addresses, developer entities, and filing timestamps to flag duplicate and speculative requests, and output a scored, exportable report a utility or regulator could use today in an active audit. No live integrations or real-time dashboard are needed yet, just a credible, defensible scoring methodology applied to data that is already public.

First Customers

The first ten customers sit inside the Texas audit itself: state regulatory staff and grid-planning teams running the current review, the law firms and consultancies already advising developers and utilities on the pause, and the handful of Texas distribution utilities fielding the bulk of large-load requests. Outside Texas, PJM transmission owners facing their own load-forecast disputes with state consumer advocates are a natural second wave, reachable through that same regulatory and legal network.

Moat Sequence

Sequence is data integrations first, starting with public grid-operator queue feeds and then utility-specific systems, then a scoring methodology validated against real audit outcomes so it can be cited by a regulator, then deep workflow embedding inside utility interconnection and regulatory affairs teams so the tool becomes the system of record rather than a one-time report.

90-Day Timeline

In the first 90 days, ship the ERCOT audit report MVP and get it in front of regulatory staff or one Texas utility ahead of the state's own audit deadline; by six months, convert that into a paid pilot with a direct queue-data integration and a second grid operator, likely PJM, in the pipeline.

Building in this space?

I've taken an agentic AI product from first build to acquisition. If you're exploring this idea or one close to it, a 30-minute call is a good place to pressure-test it.

Book a 30-min call

Sources

Texas hits pause on data center interconnections (Utility Dive) DC Hub: Interconnection Queue Texas data center pause puts US pipeline at risk of delay: BNEF - Utility Dive PUCT sets financial commitments for data center interconnection as governor pauses permit approvals - White & Case US utility large-load commitments reach 160 GW amid unprecedented PJM demand surge - Wood Mackenzie