The distinction FERC Order 2023 did not address
In July 2023, FERC Order 2023 standardized the interconnection process for generation across US ISOs and RTOs. Shared clusters. Readiness gates. Cost allocation reforms. Withdrawal penalties. The order was a long-overdue rewrite of a process that had ballooned under the weight of renewable and battery interconnection requests.
But Order 2023 was a generation order. Loads — the customers who want to plug industrial-scale demand into the grid — were not the subject of the rulemaking and are not governed by the same framework. Every US ISO handles large-load interconnection on its own terms. ERCOT runs legacy full-study plus a new Batch Study framework under SB 6. PJM has Cycle 1 under a shared-cluster reform. MISO runs DPP Phase III. SPP runs DISIS. NYISO runs annual Class Year studies. CAISO hands the process to the utilities and averages 9.2 years to commercial operation.
A developer filing a 420 MW data center across three ISOs today is filing three different packages, against three different review cycles, with three different deposit schedules and three different withdrawal penalty structures. There is no standardized cross-ISO tooling for this.
What the current tooling actually serves
The interconnection queue tooling market is real, and it is growing. GridUnity sits at the center of it: the platform of record for the majority of US ISOs and RTOs, processing generation interconnection requests from the utility side. Pearl Street Technologies (now part of Enverus) ships power flow modeling used by MISO and SPP. Nira Energy helps generation developers identify MISO POIs and pre-model system impact studies.
Each of these tools serves a legitimate, specific customer: the grid operator, the utility engineering team, or the generation developer. None of them serves the data center developer filing a large-load interconnection request. The tooling simply does not exist at the category level.
The absence is not accidental. Large-load interconnection was a small category five years ago. A handful of crypto miners and a couple of large industrial loads per year, with processes that each utility could handle manually. The product gap opened wide when the category multiplied overnight in 2024.
The customer signals pointing to an open market
In February 2026, ERCOT's large-load queue hit 226 gigawatts — four times the prior year's number. 73% of that volume came from data centers. ERCOT contracted McKinsey to help redesign its large-load interconnection process because the legacy framework cannot handle the volume.
In Virginia, Dominion filed a formal large-load queue process under the State Corporation Commission's direction. The process introduces a 100 MW threshold, a 300 MW individual cap, and batches of approximately ten projects. Microsoft and Google both filed to participate in the proceeding — a clear signal that the largest buyers see the current process as broken.
CenterPoint Energy reported a 700% year-over-year increase in large-load interconnection requests between late 2023 and late 2024 (from roughly 1 GW to 8 GW). Dominion holds 70 GW of delivery point requests against a peak historical load of 24.7 GW. ComEd reports 5 GW actively in engineering and 13 GW more in pre-engineering dialogue.
The customer demand is not hypothetical. It is a visible, structural demand signal for software that sits between the developer and the interconnection process. The category just has not been served.
The regulatory velocity problem
A generation interconnection tool that was accurate in December 2023 is still roughly accurate today. The framework is stable. Large-load tooling is the opposite: the rules are being written in real time.
FERC's December 2025 co-location order directed PJM to file three new transmission service options by February 2026. The first set of filings landed on schedule. Protests from Vistra, Constellation, and the Data Center Coalition landed in March. The reply round closes April 17. The transition deadline runs through December 2028.
Texas SB 6 is being implemented through a PUCT rulemaking (16 TAC §25.194) that went to draft in March 2026. Comments close April 17. The ERCOT Batch Study framework was filed for the February 20 Open Meeting.
Ohio's AEP tariff drove an 81% collapse in AEP's load forecast between late 2025 and early 2026 — a decline so steep that a state house bill is now attempting to extend the 85% take-or-pay structure statewide.
Virginia's new GS-5 rate class takes effect January 2027. Georgia's PSC unanimously approved a 9,885 MW Georgia Power expansion in December 2025 with mandatory cost allocation back to large loads in the next base case. Maine passed the first statewide moratorium on data centers over 20 MW through November 2027. Oklahoma SB 1488 followed.
The shared structural fact across all of this: the rules change, and the rules change in a way that affects the interconnection filing specifically. Any software that encodes the rules statically is out of date the week after it ships. The regulatory graph is a live data structure, not a feature list.
The technical shape of the product
A data center large-load interconnection workflow, when we actually decompose it, is a graph over four objects: the project (load profile, site control, financial readiness), the grid (substations, base cases, upstream upgrades), the regulatory process (ISO-specific tariff, filing requirements, study clocks), and the counterparties (utility, ISO planning staff, counsel, regulators).
Every existing tool addresses one corner of this graph. GridUnity lives inside the ISO's process. Pearl Street lives inside the grid model. Project management tools live inside the project. None integrates the full graph, and none treats the regulatory process as a first-class object that updates as rulemakings land.
A product that treats the regulatory graph as the primary object — one that ingests every filing, every protest, every ruling, and updates the customer's workspace in real time — has a structural advantage over any in-house team trying to maintain the same graph in PDFs. This is the insight that shaped Queue Intelligence.
Why incumbents will not close the gap
Three structural facts keep the opening wide. First, the predevelopment workflow is inherently multi-party and multi-jurisdictional. A hyperscaler's internal tool only needs to serve one hyperscaler's pipeline; the regulatory landscape varies across 50 states, 7 ISOs, and hundreds of utilities. Building comprehensive coverage requires serving many customers, which creates network effects that benefit a startup, not an internal tool.
Second, the established consulting and legal ecosystem profits from complexity. CBRE, JLL, and Latham & Watkins generate revenue proportional to the difficulty of the process. The law firm billing model ($1,000–$2,200 per hour) means a software tool saving 100 hours of legal work per project creates $100,000–$220,000 in value — and the customer will rationally pay $20,000–$50,000 in software fees to capture it. But the law firms themselves have no incentive to build the tool.
Third, the regulatory landscape is changing faster than internal teams can adapt. When every month brings a new FERC order, a new PUC proceeding, or a new ISO tariff amendment, a dedicated software company tracking the graph has a velocity advantage over in-house teams whose primary job is construction, financing, or operations.
What we are building
Queue Intelligence is a real-time portfolio view across every US ISO, RTO, and major utility. It encodes the process differences — cluster windows, readiness requirements, withdrawal penalties — so a project that moves across ISOs does not start from a blank form each time. It ingests ISO queue, outage, and base case data nightly and surfaces realistic headroom per candidate POI. It generates auto-populated submittal packages — load profiles, one-line diagrams, voltage tolerances, SCADA specs, and financial readiness attestations. It tracks every deadline, every deficiency letter, and every ruling that matters for your project, and pushes updates into Slack, Teams, Asana, or Linear.
Our ERCOT playbook is live. Our PJM playbook is live. Our MISO, NYISO, and SPP playbooks are live. Our regulatory graph updates within 24 hours of every FERC or PJM filing. Our pilot tier is $4,500 per project per month.
The hardest part of building a data center is the paperwork that precedes the construction. The second-hardest is reading the regulatory graph that determines whether the paperwork even belongs in the same category as the law when it was written. We have made the first part auditable and the second part live.