At Framestore, I develop internal applications the studio relies on, including review, scheduling, and messaging tools. I delivered the Public Links service end to end, with a React front end and serverless AWS back end; caching cut repeat load times from 348 ms to 21 ms.
I improved the service’s average response time from 3.3 seconds to 290 ms and set up its infrastructure as code in Terraform. I also led development of Framecast, implementing live shared cursors and drawing tools with WebRTC.
I’m the founder and sole engineer behind Garzoni, an AI-powered financial learning platform live on web, iOS, and Android. I built its Django and PostgreSQL API, React web app, and React Native mobile app, and developed an AI finance tutor that uses course-content search and voice responses.
I operate Garzoni in production with Docker, Railway, and Vercel, with automated database backups and CI checks for vulnerable dependencies and leaked secrets. Earlier, at Brixton Radio, I wrote scripts to keep social media analytics in sync and helped automate the transfer of audience statistics to Google Sheets.

