Partner Engineer at Meta. I build data pipelines, AI agents, APIs and the systems that connect them — from architecture to deployment.
A software engineer who turns ambiguous problems into systems that ship — and stay shipped.
My work sits at the intersection of large-scale platforms and the businesses that build on top of them — integrations, APIs, and the partner ecosystem that ties them together. I joined Meta as a Partner Engineer to work directly on that surface.
Before Meta I contracted in financial services at JP Morgan. The throughline is the same: regulated, high-stakes environments where the engineering has to be correct, observable, and defensible — not just clever.
Outside the day job I build and ship my own tools. Real products, public, with documentation and pricing — because the fastest way to learn a domain is to put something live in it.
Supporting technical integrations for messaging platforms operating at global scale.
Built a Spring Boot inventory microservice and Python automation pipelines across a distributed, ETL-heavy ecosystem — using Project Loom virtual threads to cut bulk data registration time by 60%.
Designed a crash-resistant Python orchestration microservice on Kubernetes, optimised for spot instances — cutting infrastructure costs by ~70% — with Helm-based IaC and asynchronous APIs.
Grounding in control, perception, and applied machine learning — the foundation behind the AI tooling and automation I build today.
Structured UK property intelligence from Rightmove — price, beds, baths, coordinates, agent details, tenure and images across buy, rent and sold. Pay-per-result, no subscription. Currency stored in pence to kill floating-point errors.
Most AliExpress scrapers return products. This one returns suppliers — deduplicated store profiles with rating, age, followers, 180-day sales and contact links. Built for sourcing teams and dropshippers who need to vet, not just browse.
Applied-AI work at the intersection of language and medicine — part of an ongoing thread of healthcare-adjacent tooling aimed at cutting clinical busywork.
Turning clinician speech into structured documentation to cut down the paper trail — speech recognition feeding a document-generation pipeline.
Open to solutions-architecture roles and senior contract engagements — especially anything touching data pipelines, automation, APIs, or applied AI.