AI infrastructure is constrained by power, not capital. We identify which sites actually have it, and produce the structured intelligence that tells investors and developers whether a site deserves commitment, before any engineering firm is called.
Five binary kill criteria. Toggle what your site has. These are the gates every AI infrastructure site must pass before any further analysis is warranted.
The AI infrastructure buildout is not limited by chips or capital. It is limited by access to power. Sites that already have it are in a category of their own.
In primary data center markets, the average wait for a new utility connection now exceeds four years. An existing active connection might not wait in that queue. And that might change the entire investment timeline.
The shift from training to inference requires proximity to end users. Centralized hyperscale campuses give way to regionally distributed sub-20MW deployments. The sites exist. They are just not qualified.
Western Canada's hydroelectric grid runs below 15g CO₂ per kWh. For capital with ESG mandates, this opens categories of investment that fossil-fuel grids cannot qualify for. These connections are structurally scarce.
Capital allocators and developers want to move fast. Most sites arrive without structured intelligence sufficient to make a commitment decision before an engineering firm is engaged. That gap is what we close.
Every site is evaluated against a structured framework developed specifically for AI infrastructure opportunities below 20MW. The goal is not to predict the future; it is to identify fatal flaws before capital, engineering resources, or management attention are committed.
Utility capacity, connection type, and the realistic path to the power level the site needs.
Proximity, transmission voltage, and what the interconnection process requires for the target load.
Thermal management feasibility, water access, and physical site characteristics for the target density.
Carrier availability, route diversity, and the cost of closing any connectivity gap identified.
Zoning status, environmental risk, and the friction level of the local permitting environment.
Skilled labour availability, competing supply, and demand signals from the operator market.
The site's ability to scale beyond Phase 1 in terms of land, power headroom, and structural constraints.
Every data point in a Potentia qualification output carries a source and a confidence tag. Nothing is presented as fact without a traceable origin.
This tagging system means every report is explicit about what is known, what is estimated, and what still needs an engineer. That transparency is the product, not a caveat.
Walk through the methodology with us →The report is the product. Enter your email, and we will send you the illustrative case and offer a call to walk through it if you want.
An illustrative case study was produced on a publicly available industrial site in the Dallas-Fort Worth Garland corridor. Every section of the methodology is applied in full. No real client data. Full analytical process visible.
We will send the full illustrative case to your inbox. If you want to walk through the methodology or discuss how it applies to a specific site, we will offer a call at the same time. No obligation.
How a qualification process unfolds from initial screen to qualified outcome. Site details anonymized.
Five binary criteria evaluated against public records and utility databases. Active power connection confirmed from the utility rate schedule. Industrial zoning confirmed from municipal records. Flood zone status is clear from provincial mapping.
Existing 8MW connection confirmed from the primary utility filing. Substation proximity mapped via GIS, confirmed within 3.2 miles. Three-phase power scaling pathway modelled to 20MW based on available transmission capacity at the serving substation.
Two carrier routes were identified within 1.4km of the site boundary from public network mapping. Route diversity confirmed as adequate for AI inference workloads. Direct carrier confirmation is flagged as the required next step before capital commitment.
Primary risk identified: cooling retrofit cost for AI rack density. Secondary risk: workforce access in the sub-region. Both risks were modelled with directional capital impact ranges and specific engineering confirmation steps defined.
Not a consulting firm. Not a broker. Not an engineering company. An intelligence layer built for the moment in the decision process that everyone skips over.
The buildout of AI infrastructure is accelerating faster than the grid can respond. New utility connections take years. The sites that already have power, former industrial campuses, legacy manufacturing facilities, cannabis operators, and mill sites, have become structurally valuable assets. Most of them do not know it yet.
The standard process puts engineering diligence after the LOI. By then, capital is already committed, and time has been spent on a site that may not qualify. The screening step that should happen first, before any of that, is either done manually on spreadsheets or skipped entirely. That is the gap we built for.
We produce structured site qualification intelligence that tells capital allocators, developers, and brokers whether a site deserves deeper commitment — before any engineering firm is called, and before any capital is moved. We are not the next step in the process. We are the gate before it.
The methodology has been reviewed and stress-tested in conversation with active practitioners across the infrastructure stack, prior to any commercial engagement.
Practitioners with direct experience in large load interconnection, rate structure, and utility design queue processes in Canadian and US markets.
Operators and deployment leads with direct experience qualifying and building sub-hyperscale AI infrastructure in enterprise and neocloud environments.
Infrastructure investors and capital allocators evaluating sub-20MW AI site opportunities across Western Canada and Ontario deal flow.
Developers and acquisition teams with active pipelines of industrial sites across the 1MW to 20MW range in Western Canada and select US markets.
Three levels designed to give the right answer at the right moment in the decision process, all faster and at a fraction of the cost of traditional screening.
A fast binary read against the core kill criteria. Designed to eliminate dead-ends quickly, before any engineering engagement, before any capital is moved.
The complete site qualification. A structured intelligence brief that tells investors, developers, and realtors exactly what a site can support and what stands in the way.
Ongoing monitoring for developers and capital partners managing active site pipelines. Grid capacity tracking, regulatory changes, and deal flow intelligence across a target geography.
Existing hydroelectric connections carry a structural scarcity premium. Provincial allocation constraints on new large industrial loads make active connections on former industrial sites among the most defensible power assets on the continent for ESG-mandated capital.
Active deal flow across Ontario's industrial corridor. Strong fiber infrastructure and an established data center market with favourable regulatory conditions for new industrial power users.
ERCOT's deregulated market offers competitive power pricing for AI inference. The Pacific Northwest mirrors the hydroelectric profile of Western Canada with established hyperscaler demand clusters already in place.
Start with a 30-minute conversation. Bring the site location, a sense of the power connection, and your questions. We will tell you honestly whether it is worth pursuing further.