The clinical fleet is where the platform began and where the evidence runs deepest. MammoChat launched as a breast health companion for patients navigating a new diagnosis, grew into a live NIH-funded clinical trial, and proved that governed AI could work in a clinical setting without hallucinating, hedging, or hiding its sources. The four products that followed inherit the same constraints from Medicine: every response cites guidelines, every session is ledgered, and every encounter compounds the community intelligence.

The same governance triad that keeps MammoChat honest also manages a real estate portfolio across Lake Nona and London. REALTY is the property intelligence layer, drawing from public records and governed listings. RUNNER handles operations for every property in the portfolio. NONA connects buyers to realtors in the Lake Nona market. The domains are different, but the compliance architecture is identical: every interaction ledgered, every claim sourced, every transaction governed.

FinChat and LawChat prove the thesis that governed AI is not a healthcare-specific idea. Each product inherits its industry’s constraints from the Machine fleet: securities disclaimers for finance, unauthorized-practice-of-law guardrails for legal. The compliance requirements change, but the architecture that enforces them does not.

GALAXY is the internal operating surface where governors and developers interact with the full governance graph. It is the control plane for everything in this library: service contracts, coverage audits, learning patterns, and deployment status. GALAXY is currently in development and will be the first product in the fleet that governs the fleet itself.