MOVE

• Business Overview & Market Opportunity

  • Corvex builds and operates AI infrastructure for AI training and production inference workloads, with three core product lines: integrated AI factory GPU clusters, the Corvex Token Factory open-weight inference API platform, and patent-pending confidential computing for AI model IP security.
  • Key demand drivers: 1) AI factory: Global demand for energized, permitted AI compute capacity far outpaces supply, with power availability as the core binding constraint; 2) Corvex Token Factory: Inference now represents the majority of AI workloads, and a growing underappreciated shift of routine workloads from expensive closed-weight models to lower-cost capable open-weight models is creating new unmet demand; 3) Confidential computing: Growing awareness of AI model IP security risks and requirements for regulated enterprises creates a large unaddressed market opportunity.
  • Customer choice drivers: AI factory customers prioritize Corvex's speed to capacity (ability to modify and deliver configured clusters far faster than industry standard timelines) and differentiated security capabilities that appeal to regulated and security-conscious customers.

• Business Model & Operating Metric Framework

  • All current AI factory revenue comes from fixed-term, taker-pay contracts: customers pay a fixed reservation fee regardless of utilization, delivering stable contracted revenue rather than volatile spot market revenue.
  • All new cluster deployments are required to clear a project-level return hurdle (accounting for power, hardware, financing, contract terms, residual value, and counterparty credit) before approval; deals that do not meet the threshold are not signed.
  • Management has outlined core operating metrics for investors to track progress: contracted power capacity, contracted annualized recurring revenue on live compute, and total brought-in-service capacity. The $3.8 million Q2 revenue only reflects capacity that went live in the quarter and does not reflect the full level of underlying business activity or total contracted volume.

• Operational & Leadership Updates

  • Corvex Token Factory v1 is now live in closed alpha, with additional releases planned for Q3 and Q4 2026. Planning for v2 of Corvex's cloud management software is complete, and development execution is underway to improve platform automation, reliability, and scalability.
  • Significant progress was made on securing additional power capacity for future expansion, with more details expected to be shared in the near term. Four new experienced leaders and board members were added during and after the quarter to build out core scaling capabilities: Michael Craig (VP of Architecture and Site Operations, 25+ years of large-scale infrastructure experience), CFO Chance Moreland (experienced infrastructure capital markets financier), and new board members Nick D'Onofrio (44-year IBM veteran of semiconductor and server development) and Patrick Fleury (CFO of TerraWolf, $8B+ data center financing experience) to strengthen governance and strategy.
  • Post-quarter capital structure simplification was completed: all Series A, B, and C preferred stock were converted to common stock, leaving only Series D preferred stock outstanding, for a total of ~56.6 million as-converted common shares outstanding. A resale registration statement for 53.4 million existing shares was filed, with no new primary offering or proceeds to Corvex.

Management did not provide explicit full-year financial guidance, but outlined forward-looking expectations and priorities for scaling:

• Management expects predictable, recurring non-cash stock-based compensation expense of a similar large magnitude to Q2 2026 to be recorded in each quarter through 2030, stemming from equity awards issued in the merger.
• Scaling the AI infrastructure business will require ongoing additional capital; funding needs will vary based on the pace of infrastructure expansion. Recent Blackwell GPU expansion was funded via a mix of customer prepayments, debt financing, and existing cash.
• Work to remediate the existing material weaknesses in internal financial controls is underway, with risk assessment and scoping targeted for completion in the current fiscal quarter, followed by control development and testing. No specific remediation timeline was provided.
• Corvex plans to expand public reporting on its core operating metrics (contracted power, live contracted ARR, in-service capacity) as these figures grow large enough to be meaningfully disclosed.

This is Corvex's first full reporting period after completing its merger with Corvex Legacy Holdings on March 19, 2026, including the AI cloud computing business (prior periods only included the legacy healthcare business, which was fully disposed of in Q2 2026). All $3.8 million of Q2 2026 total revenue came from the AI platform and services segment. Total six-month revenue for the AI business from March 19 to June 30 was $4.3 million. As of August 2026, contracted annualized recurring revenue (ARR) on live, customer-accepted revenue-generating compute capacity is approximately $22 million. AI platform cost of revenue (excluding depreciation and amortization) was $2.1 million for the quarter, with depreciation and amortization of $2.7 million. Corvex has three planned product segments: 1) AI Factories and GPU Clusters (the only segment currently generating revenue, 100% of Q2 revenue), 2) Corvex Token Factory (inference platform, currently live in closed alpha, no revenue yet), 3) Confidential Computing (offered as a differentiator for the first two segments, planned as a future standalone offering, no revenue yet).

• Unremediated material weaknesses in internal controls over financial reporting remain in place, with significant work required to achieve Sarbanes-Oxley compliance, even though a remediation plan is underway.
  • Revenue is currently concentrated in a small number of customers, creating counterparty credit and revenue concentration risk.
  • The pace of business scaling is dependent on successfully securing additional permitted power capacity, which is a binding industry-wide constraint that could slow growth.
  • Forward-looking results depend on successfully raising additional capital to fund infrastructure expansion; the ability to secure capital on favorable terms is not guaranteed.
  • Corvex Token Factory is still in early closed alpha development, and requires rapid iteration and product growth to achieve market fit and revenue contribution.
  • Contracted ARR may see transitory dips in future periods if Corvex re-purposes reserved capacity to Corvex Token Factory inference workloads during contract renewals.

There was no question and answer section held during this conference call.