Nuclear Restart Actuarial Intelligence Service

Impossibility Negation: 'You cannot actuarially price nuclear restart risk because there's no historical loss data for reactor restarts after decades of dormancy.' AI can now synthesize reactor condition data, decommissioning records, regulatory history, and seismic changes into a restart risk model without historical restart-specific loss data.

Markel Insurance white paper confirmed actuarial data gap for nuclear. Newfront Insurance confirmed. ANI and NEIL confirmed as primary nuclear insurance pools. DOE $1.52B loan guarantee for Palisades confirmed. 15 reactors coming online globally in 2026 confirmed from IAEA data and Carbon Credits reporting.

Expanded from restarts-only to 'actuarial intelligence for nuclear energy transitions' covering SMR deployments, decommissioning, and data center co-location insurance. Combined TAM $50-100M growing with SMR market.

43% conviction. Survived Pass 1 with buyer concentration worry (5-10 nuclear insurance entities globally). Inter-industry gap confirmed (nuclear engineering × insurance actuarial). Regulatory tailwind from nuclear renaissance.

TAM of $5M-15M initially, growing to $50-100M with SMR market — single-source estimation from Pass 1. Price-Anderson Act structure is REGULATED, limiting what a data platform can influence. Liability tiers set by law, not market forces.

Newfront Insurance (nuclear broker, not data platform but fills advisory gap). Nuclear engineering consultancies (AECOM, Bechtel, Burns & McDonnell) provide bespoke nuclear risk assessments at $500K-2M. No startup doing standardized actuarial intelligence for nuclear restarts specifically.

structural barrier GATE FAILED: Bespoke nuclear risk assessment is intentionally bespoke — each reactor is genuinely unique (PWR vs. BWR vs. SMR, different shutdown duration, different site geology). A standardized platform producing artificially comparable scores for genuinely different risks would be less accurate than custom assessment. Nuclear insurers CHOOSE manual bespoke assessment because the situation demands it. DISTRIBUTION GATE FAILED: Fewer than 15 potential customers globally. Cannot reach 100 paying customers because fewer than 100 potential customers exist in the entire world.

0% conviction. Two gate failures: (1) Bespoke-by-necessity barrier — standardized platform would be less accurate than custom assessment for genuinely unique reactors; (2) Distribution — fewer than 15 potential customers globally. Nuclear renaissance is real, actuarial gap is real, but market structure makes this a consulting business to a cartel, not a scalable startup.