Work
These aren’t products for sale. They’re evidence — of the kind of modern, scalable, ML-heavy systems I design, build and lead. Each one stands behind one of the services I offer.
Codamor
Turns a coding pipeline into a service-delivery continuous-improvement loop. A requirement becomes a governed, auditable system — security, architecture, UX, testing and operations applied as AI perspectives, models from different vendors checking each other’s work — and the loop keeps going, watching the live service and feeding the gaps back in.
Plenio
EU AI Act compliance middleware. Puts a named human back into the loop at the moment AI output influences a decision, and records that moment as immutable, audit-defensible evidence.
Intuitra
A production ML-driven customer engagement platform with sophisticated orchestration. RAG-ML chat is just the customer-facing surface — over an authoring CMS, multi-format ingestion, Voice-of-Customer analytics and codeless workflow automation.
Service transformation — US fixed-wireless ISP
Diagnosed knowledge and service gaps across a national operator: 5,560 calls analysed, a cost model of the real call drivers, a ranked quick-win backlog and a service maturity model to steer the programme.