HADLEYLAB
Hadley Lab — Clinical AI Research & Governed Intelligence
SERVICE CONTRACT · VIEW: GOV
HADLEYLAB · SERVICE CONTRACT · CANONIC ∩
Nine AI products for clinical medicine, genomics, real estate, law, and finance, each governed to 255 before it reaches a single user. MammoChat is the flagship: free for every patient, funded by a $2M state grant, running as a live clinical trial. Community learning from over 20,000 real encounters compounds intelligence across the entire fleet, so every question asked makes every product smarter.
The clinical fleet serves breast health, oncology, genomics, preventive medicine, and Caribbean cancer navigation, with each product inheriting its constraints from Medicine. MammoChat is the flagship: free for every patient, funded by a $2M state grant, and running as a live clinical trial (NCT06604078). Every response cites its guideline. Every product is validated to 255 before deployment. Every encounter compounds the community intelligence that the rest of the fleet draws from.
When a patient in Port of Spain asks where to get screened, the answer enters a governed ledger. When a clinician in San Francisco corrects a staging recommendation, the correction compounds across the fleet. This is not crowd-sourced AI; it is append-only, evidence-tagged, privacy-governed intelligence that grows with every encounter and never forgets what it has learned.
PMWC in Santa Clara, Bamberg Health in Miami, CAOH in Port of Spain, and the Trinidad launch in July 2026. Each event is a chance to prove the platform in front of clinicians, researchers, and policymakers who will use it.
The same governance triad that keeps MammoChat honest also manages a real estate portfolio spanning Lake Nona and London. REALTY provides property intelligence from public records, RUNNER handles listing operations, and NONA matches buyers with realtors in real time. Different domain, identical compliance architecture.
FinChat and LawChat extend the fleet into professional services, where the compliance requirements differ but the architecture that enforces them does not. Each product inherits its industry’s constraints from the Machine fleet, so securities disclaimers and unauthorized-practice-of-law guardrails are composed the same way clinical guidelines are.
Everything published by Hadley Lab, organized by what you came to read. The essays chronicle the building of governed AI in real time. The papers present the clinical evidence. The books compile the theory. Start anywhere; each piece stands alone, and every claim is sourced.
Every AI on this site must score 255 out of 255 before it reaches a single user. Eight governance dimensions, each scored as a binary pass or fail. The compiler validates every scope on every commit, and if any dimension drops below the threshold, the product does not deploy.
Declaration, Evidence, History, Community, Practice, Structure, Learning, Language. The compiler checks all eight on every commit. No exceptions, no waivers, no manual overrides.
Hadley Lab is the first organization certified to MAGIC 255, and the proof that the platform works at production scale. A clinical research lab that governs five clinical AIs, a $12.2M real estate portfolio, and professional intelligence services under the same compliance engine. Fourteen years, 7,062 commits, 6 patent families, 65 peer-reviewed publications, and over $38M in funded research, all under one governance standard.
Hadley Lab is the proof; the Machine fleet is the specification that makes it possible. The Foundation certifies organizations, Industries constrain domains, and MAGIC operates the governance engine. Together they compose the compliance architecture that every product in this library inherits.