Artificial Intelligence
Applied AI companies with deployed models, proprietary data rights and contracted revenue.
How We Read This Sector
Artificial intelligence is the sector we are asked about most often, and the one where the gap between narrative and substance is widest. A credible public-market AI business is not defined by the sophistication of its models but by what it owns, what it has deployed, and what customers have contracted to pay for.
What We Examine
Five areas that carry disproportionate weight in this sector.
Model and data ownership
Whether the company owns its models and weights outright, and whether its training data was obtained under rights that survive a change of control and public disclosure.
Dependency on third-party foundation models
Companies built entirely on a single external model provider carry a concentration risk that public investors examine closely. We look at what would remain if that provider changed terms.
Deployment evidence
Production deployments, named customers, usage volumes and renewal history — distinct from pilots, letters of intent and proofs of concept.
Unit economics of inference
Gross margin after compute cost, and how that margin behaves as usage scales. Many AI businesses report revenue growth alongside deteriorating margins.
Regulatory exposure
Sector-specific obligations where models touch health, credit, employment or biometric data, and how the company documents compliance.
Where Companies in This Sector Get Caught
AI companies frequently arrive with strong revenue growth and weak corporate foundations — intellectual property assigned informally, contractor agreements without IP transfer clauses, and training data whose provenance cannot be evidenced. These are the issues that most often stall an RTO process during due diligence, and they are far cheaper to correct before a process begins.
Sector Fit Is the Starting Point, Not the Test
Beyond sector-specific factors, every company is assessed against the same conditions: a verifiable asset, documented ownership, auditable financial reporting, management capable of operating a public company, and a budget to fund preparation.
Also in Software, AI & Data
Machine Learning
ML infrastructure, MLOps and applied modelling businesses with a defensible technical position.
What we reviewEnterprise Software
B2B SaaS and enterprise platforms with contracted recurring revenue and measurable retention.
What we reviewDeepTech
Science-led companies with granted patents, laboratory validation and a defined commercialisation path.
What we reviewBuilding a Artificial Intelligence Company?
Submit your company for a confidential preliminary assessment. Submission does not guarantee selection, financing, a transaction or a public listing.
Confidential assessment · Selective admission · No obligation