What I've shipped, when, and where.
Nearly a decade building software for safety-critical industrial systems — biometric algorithms, real-time pipelines, large-scale data platforms — then data-team leadership, now enterprise AI. Same engineering rigour, pointed at agentic ITSM and AI governance.
For the tools and the game-style skill tree, see Skills.
As AI Strategy Lead, designing agentic ServiceNow tooling and an AI Management System (AIMS / ISO 42001) for enterprise service management — turning ambiguous problems into governed, shipped systems, and building Xcession's AI practice as a differentiated capability.
A studio of side products shipped solo, full-stack — CoachPaper (grassroots football), Nose Codes (fragrance discovery), Man Buys Jewellery — plus warrghh.com itself. Design, build and deploy, end to end, on one reusable house stack.
As Data Team Lead, an end-to-end ML pipeline for intelligent scheduling optimisation, built on MLflow and Spark — tracked, reproducible and retrainable.
As Data Architect, data systems for a global chemical-ingredients supplier — ingestion, modelling and warehousing.
As Data Team Lead, architected the platform end to end and drove the legacy migration to full GDPR compliance. Recruited and led a team of 10 across algorithms, back-end and embedded, owning the full SDLC and review process.
As Senior Software Engineer, built the biometric safety algorithm suite — real-time core-body-temperature, mental-fatigue and device-fit prediction — reducing critical safety incidents across participating industrial sites.
A production-grade calibration system for temperature sensors at manufacturing end-of-line — custom hardware, application software and a calibration model — for a ~10% improvement in sensor accuracy.
My first engineering role: a neural-network person-down / fall-detection system analysing accelerometer and gyroscope data for real-time detection in high-risk environments — 92% accuracy, with MLflow for tracking.
An automated test-collection framework for REST API validation: Postman for case management, PyTest with Django for execution. Cut manual testing and caught regressions earlier in the pipeline.
