Regulated sector experience
Head of Commercial & IT at a UK clinical diagnostics company — NHS-facing, regulated, high-stakes.
Not adjacent to healthcare — inside it. Responsible for commercial strategy, IT systems, vendor relationships, and the interface between laboratory operations and technology procurement. Every decision existed within NHS procurement rules, clinical governance constraints, and ISO 15189-adjacent quality frameworks.
- Commercial leadership: NHS trust-level negotiations, supplier contract management, tender strategy, and commercial pipeline oversight
- Technology procurement: evaluated and procured digital pathology platforms — Indica Labs HALO, Deciphex, Leica Biosystems, 3DHISTECH, Paige AI
- HL7 integration: operated and troubleshot within HL7-native LIMS environments where data integrity had patient-level consequences
- NGS pipeline context: involved in next-generation sequencing pipeline development at a commercial diagnostic lab
- ISO 15189-adjacent governance: validated workflows, documented change rationale, managed audit trails — compliance as operational discipline
Risk-preserving AI transformation
AI that doesn't quietly do the wrong thing when the stakes are high.
The architecture pattern I apply across all AI systems comes directly from regulated-sector thinking: human gates on destructive steps, full audit trails, fail-safe rather than fail-silent, and canary rollback before full exposure. AI that removes these controls isn't more capable — it's more dangerous.
- T-V2 canary controller: production traffic stays at v1_weight=100 until a human sets
v2_api_ready=true— no automatic production cutover - Quality gate: API errors escalate to human review, never silently pass — same discipline as clinical validation
- Rights screening with full audit trail: every Block/Escalate/Pass decision is recorded with the matching rule
- Worthington review queue: AI generates, human reviews, system auto-publishes only after explicit approval or hold expiry
- DPAPI-encrypted credentials vault: no plaintext secrets in automation paths — same principle as patient data access control
Marketplace strategy
Lab Locum Marketplace — built from real sector knowledge, not desk research.
Building a UK marketplace for AP/BMS (Anatomical Pathology / Biomedical Science) and histopathology consultants. The opportunity is real because I spent years inside the sector understanding the workforce gap, professional registration requirements, and procurement pain from both sides of the table.
- UK healthcare workforce gap: AP/BMS and histopathology consultant locum demand consistently outstrips available supply
- Existing contacts from years in cellular pathology commercial work — not a cold-start marketplace
- Regulatory awareness: HCPC registration requirements, lab director sign-off, clinical governance obligations
- Commercial model grounded in understanding of how NHS trusts and independent labs actually procure locum resource