Initially hired as a Lead Dev, I decided to step down from that more managerial role to lead the development of a cross-domain cookie consent service which I designed, built the proof-of- concept (Python+Flask, JS, Heroku), then automated CI/CD with Github Actions and Pulumi, deploying Docker images to Google Cloud Platform. I brought the service up to production readiness and liaised with other government departments to begin deploying it across the GOV.UK estate.I was also seconded to the GOV.UK AI team to rapidly rearchitect an AI Chatbot service to mitigate security risks, fix a serious stability issue, and improve performance in time for the AI Safety Summit, where the Prime Minister discussed the service with Elon Musk.

I was a founding member of the team to build a tool for publishing the post-Brexit UK trade tariff. Itranslated a Ruby-on-Rails prototype to a Python+Django web application UI and REST API;designed and optimised a bi-temporal database modelling a complex hierarchical tree structure;setup Github Actions for continuous deployment; designed and implemented XML import and exportof multi-gigabyte data files. I introduced documenting architecture with ADRs and wrotedocumentation with Sphinx. We were able to deliver the MVP in time for Brexiton the 1st Jan 2021, publish tariff changes to HMRC during the year, and prepare for changes tothe WTO Harmonized System which took effect from 1st Jan 2022.

I developed the MOJ Analytical Platform, a PaaS providing a data lake, development environments and hosting for data science applications. I used Terraform to setup a Kubernetes cluster on AWS; setup continuous integration and deployment; wrote Go, Bash, Python and R; built a Python+Django REST API, a Node JS web app and an OIDC authentication proxy, and AWS Lambda functions in Python and Javascript; and integrated with ElasticSearch to filter ELK stack application logs.I directly saved the project £350k per year by implementing an authentication proxy for the R Studio application.

I worked on the Government As A Platform (GaaP) team, on an internal tool (Python+Flask, PostgreSQL) hosted on GOV.UK PaaS, for annual performance reviews.This project pivoted to providing a federated login service for GaaP products, allowing users to authenticate with their existing credentials.I built a prototype using Python+Flask implementing OpenID Connect (OIDC) hosted on AWS using Kubernetes and Docker.

I worked on the Civil Legal Aid service, a case handling system in Python+Django with Node JS for SocketIO and an Angular JS UI used by call-centre staff; a public-facing website in Python+Flask; and a Django REST API. The service was delivered in time for the handover from an external supplier whose contract was ending.

I line-managed and mentored a distributed team, guided the development and launch of a new product (Python+Django, PostgreSQL, Erlang) and led the internationalisation of the site and launches into several countries.I also introduced Git, Agile, code reviews and a standard development environment using Vagrant.