Keel Infrastructure Corp. (KEEL) Company Overview

What does Keel Infrastructure do?

Keel Infrastructure Corp. is a Nasdaq- and TSX-listed digital-infrastructure developer trading under KEEL. The business is being rebuilt around a scarcity thesis: AI and high-performance-computing customers need reliable power, suitable land, connectivity, and credible permitting and construction paths. Keel aims to assemble those inputs into lease-ready campuses in Pennsylvania, Washington, and Québec. Its official portfolio overview describes a North American platform rather than a conventional colocation operator with a mature base of contracted tenants.

That distinction is essential. Keel’s reported revenue still comes mainly from legacy Bitcoin mining, hosting, electrical services, and energy sales. The target model develops powered sites, secures long-duration customers, and converts projects into infrastructure cash flows. Analysis must therefore separate declining legacy economics from the unproven economics of leased AI infrastructure.

How should the company be classified today?

How does Keel make money during the transition?

Bitcoin mining contributes revenue when Keel’s machines supply hashpower to a pool; economics depend on Bitcoin price, network difficulty, efficiency, and electricity cost. Energy sales come mainly from Stronghold’s power assets, while hosting and electrical services are smaller. The intended model replaces volatile block-reward exposure with contracted capacity payments from data-center customers, but the Q1 2026 Form 10-Q shows that this replacement had not yet occurred.

Which revenue source still dominates?

What does the latest reported quarter show?

The first-quarter 2026 results release, shows a costly transition. Mining economics weakened while acquired power assets increased depreciation and infrastructure expense. Keel mined 388 Bitcoin in Q1 2026, versus 492 a year earlier. Its average electricity cost rose to $0.058 per kWh from $0.047, while network difficulty increased 27% year over year. Those variables turned a moderate revenue decline into a much sharper gross-profit deterioration.

Why did profitability deteriorate so sharply?

What does the annual baseline add?

Operating performance should be separated from accounting volatility. Q1 included a $41.4M loss from remeasuring digital assets and a $21.6M debt-extinguishment loss. Those items matter economically, but they are not the same as recurring data-center operating margins. The more durable concern is that gross loss, negative adjusted EBITDA, and cash burn all appeared before major campus construction began.

How did a Bitcoin miner become an AI-infrastructure developer?

The transition is an asset-conversion strategy, not a clean-sheet startup. Mining required power contracts, substations, data halls, and electrical deployment. Management is reusing those capabilities while seeking long-duration infrastructure contracts.

1. 2017
   Bitfarms was founded as a Bitcoin miner. The model built experience in power procurement, machine deployment, and high-density loads.
2. 2024
   The fourth Bitcoin halving compressed mining rewards. Industry economics became more sensitive to power price, efficiency, balance-sheet strength, and access to alternative uses for energized sites.
3. Mar. 2025
   The Stronghold acquisition closed. It added two Pennsylvania waste-coal power facilities with 473 MW of combined generation capacity, expanding both the power base and regulatory complexity.
4. Oct. 2025
   A convertible-note financing strengthened liquidity. The company raised capital intended to support the HPC and AI pivot, long-lead equipment, and site development.
5. Apr. 2026
   The company redomiciled to Delaware and became Keel Infrastructure. The rebrand and U.S. redomiciliation aligned the corporate identity with the infrastructure strategy and shifted SEC reporting to a domestic-issuer framework.
6. Jun. 2026
   Keel closed another convertible-note offering. The financing increased development flexibility but also raised future interest, refinancing, and potential dilution considerations.
7. Jul. 2026
   Commercial execution moved to the foreground. Ganesh Aiyer joined as President to lead pipeline expansion, and the Sherbrooke conversion advanced through municipal approvals.

What did the strategic pivot actually change?

It changed the unit of analysis. A miner is judged by hash rate, joules per terahash, Bitcoin production, and power cost. A data-center developer is judged by secured megawatts, lease credit quality, rent per kilowatt, construction cost, ready-for-service timing, power usage effectiveness, and stabilized returns. Keel still reports the first set because it has not yet earned meaningful revenue from the second.

Why are power, interconnections, and site control Keel’s strategic assets?

AI infrastructure is constrained by deliverable power on a usable schedule. A nominal megawatt is insufficient without interconnection, transmission, land, zoning, environmental clearance, equipment, cooling, and fiber. Keel’s opportunity is to shorten that chain for customers that otherwise face multi-year utility queues.

Which campuses matter most?

The Q1 2026 investor presentation indicates estimated power usage effectiveness of roughly 1.15 to 1.35 across the portfolio. That range is useful for preliminary engineering, but a DCF should not treat it as stabilized operating performance until designs, tenants, and service conditions are finalized.

Why is Sherbrooke a revealing test case?

In July 2026, the city authorized agreements to transfer and operate 96 MW of existing Hydro-Sherbrooke capacity at a consolidated campus, subject to Québec government review and other conditions. The Sherbrooke update is strategically important because it tests whether Keel can convert legacy power rights into a higher-value use without requesting incremental utility capacity.

Who competes with Keel, and what is its actual moat?

Keel faces established data-center platforms such as Digital Realty and Equinix, which bring customer relationships, operating records, and procurement scale. A second group—including Applied Digital, Core Scientific, IREN, Hut 8, TeraWulf, Cipher Mining, and other power-rich operators—is also converting mining or energy assets toward AI workloads. Keel’s own investor materials use many of these companies as relevant market peers, although their business mixes and contract maturity differ.

What could become a durable advantage?

Keel’s strongest resource is not the old mining fleet; it is the combination of power access, interconnection work, land, energy-market knowledge, and sites that can be advanced before customers commit. These assets may create a time-to-power advantage. The Pennsylvania portfolio also offers optionality across grid supply, on-site generation, and expansion acreage. Capital also permits earlier equipment orders and permitting work.

Where is the moat still unproven?

A defensible moat requires bankable leases at returns above the cost of capital. Keel still needs to demonstrate tenant credit, construction discipline, uptime, and repeatable project economics. Its current losses also weaken bargaining power if capital markets tighten. In resource-based terms, power positions may be valuable and scarce, but they become durable only when the company can organize them into contracted, financeable, operating assets.

How financially strong is Keel after the 2026 financings?

Keel entered the transition with liquidity. At March 31, 2026, it held $357.3M of cash and $168.5M of digital assets, including 2,469 Bitcoin. Working capital was $515.7M. That supports site work, but large campuses still require disciplined project financing.

What does the balance sheet say?

How did the June financing change the picture?

The June 2026 convertible-note closing materially increased gross financing capacity after the Q1 balance-sheet date. It also added fixed interest obligations and potential share dilution. A sound analysis should therefore avoid treating Q1 debt and cash as the current pro forma position, while also avoiding an unsupported pro forma total that ignores capped-call costs, spending, and post-quarter cash movements.

Who owns Keel stock, and how is it governed?

Keel has a single class of common stock with one vote per share and no cumulative voting. The redomiciliation materials indicated that, as of February 13, 2026, the company did not know of any person owning more than 10% of the outstanding shares. Directors and executive officers as a group had voting rights over approximately 1.9 million shares, or 0.31%. Strategic accountability therefore rests with the board and a dispersed shareholder base.

Why does the investor base matter during a pivot?

A dispersed register can reward milestone evidence but may be less patient with slippage. The redomiciliation proxy circular also shows that U.S. domestic-issuer reporting and Nasdaq governance now frame accountability. Keel’s addition to the Russell 3000 Index in June 2026 may broaden passive and benchmark-aware ownership, increasing attention to liquidity, dilution, and execution milestones.

What opportunities and risks could change the story?

What are the strongest growth opportunities?

The largest opportunity is to monetize the gap between AI demand and available grid capacity. Panther Creek and Sharon could establish a repeatable Pennsylvania development platform, while Scrubgrass offers longer-dated scale. Moses Lake can demonstrate a smaller, modular deployment. Sherbrooke can prove that an existing mining footprint can be consolidated into a data-center campus. Customer diversification across hyperscalers, neoclouds, enterprises, and government users could also reduce concentration.

Which risks are most material?

The filings warn that the HPC and AI strategy may not become profitable and may divert resources from mining. Development is capital-intensive, customers may delay or default, and better-capitalized operators can compete for sites and tenants. Equipment shortages, power-market constraints, cyber incidents, construction overruns, and environmental obligations can also impair returns. Stronghold’s waste-coal generation adds plant, fuel, mine-safety, reclamation, and emissions exposure that a conventional data-center developer may not carry.

Customer concentration could become a double-edged sword. One large lease can validate a campus and improve financing, but it can also concentrate credit and renewal risk. The same is already visible in legacy operations: one mining-pool operator generated 88% of FY2025 revenue. For researchers, the key question is not whether AI demand is strong in general; it is whether Keel can convert specific sites into contracts whose risk-adjusted returns exceed the cost of capital.

Which KPIs and valuation drivers matter most?

A conventional revenue-growth DCF is poorly suited because the historical base is shrinking while the target business is pre-revenue. The model should separate legacy cash flows, corporate overhead, development spending, project-level construction, and stabilized lease economics. It should also model share dilution from convertible notes and equity compensation rather than valuing the enterprise on an undiluted share count.

What should a DCF model track?

What should researchers monitor next?

The most informative next disclosure would be a signed lease revealing customer quality, megawatts, term, economics, and delivery obligations. After that, financing and construction can be assessed rather than inferred from site potential. Until then, terminal value should be scenario-weighted.

What is the key takeaway on Keel Infrastructure?

Keel is important because it represents a broader industry shift: energy-intensive computing assets are being repurposed for AI at a time when deliverable power is scarce. The company has real ingredients—North American sites, energized infrastructure, a sizable pipeline, development liquidity, and management focused on commercialization. It also has a difficult starting point: negative legacy margins, significant cash burn, convertible debt, regulatory complexity, and no mature lease portfolio visible in the latest quarter.