I build governed AI systems, orchestration layers, and production AI infrastructure for regulated and enterprise environments.
My work focuses on deterministic control layers around probabilistic AI systems, combining workflow orchestration, validation, auditability, evidence traceability, and operational reliability.
Most current work is private due to commercial, legal, and enterprise confidentiality requirements.
Public repositories demonstrate selected architecture, orchestration, evaluation, and infrastructure patterns from active production systems.
- AI execution and orchestration systems
- Multi model routing and provider abstraction
- Evidence intelligence pipelines
- Agent governance and validation layers
- Workflow integrated AI systems
- Structured reasoning and assertion analysis
- Probantum
- LawBow
Evidence intelligence and workflow orchestration platform exploring structured assertions, litigation workflows, contradiction analysis, and traceable AI assisted review systems.
Deterministic execution layer for multi provider AI orchestration with explicit separation between execution, policy, routing, and governance.
Public product and positioning layer for Probantum, a practitioner facing evidence intelligence platform.
- TypeScript
- Next.js
- AI orchestration
- Multi provider routing
- Validation systems
- Structured extraction pipelines
- Human in the loop workflows
- Auditability and traceability
- Production AI operations
30+ years across SaaS, cloud, telecoms, infrastructure, enterprise systems, and AI driven platforms.
Selected work includes:
- KPMG
- Amazon Locker infrastructure
- NHS messaging and mobility systems
- DEC Tsunami Appeal platform
- Early ISP and SaaS infrastructure through London Web
AI systems become commercially valuable when reliability, governance, workflow integration, and operational control are treated as first class architectural concerns.