One orchestrator, fifteen AI attributes
A modular, event-driven pipeline that turns free-text profiles and reviews into structured, AI-generated attributes, where adding a new one means writing one module.
Open to senior and lead backend roles, UK remote or Norwich/London hybrid
I’m a backend engineer with over a decade of experience, mostly in PHP and Laravel, building event-driven systems that move data reliably between services. Lately that has meant putting LLMs into production features in a way that stays testable and affordable. I like owning a problem end to end, from the architecture to the deploy, and I mentor the people around me while I do it.
A modular, event-driven pipeline that turns free-text profiles and reviews into structured, AI-generated attributes, where adding a new one means writing one module.
Moving a global set of services from legacy polling to event-driven messaging on AWS SQS and SNS, and onto Kubernetes.
A Python renderer that simulates embroidered lettering stitch by stitch, thread by thread, and composites it onto real garment photos.
Built a modular orchestrator that generates AI-written tutor profile content from one versioned attribute store. Each module declares its inputs and output schema, so adding a new attribute means adding one module, and domain events trigger re-generation of only the attributes that are affected.
Built a server-driven end-of-tuition wizard where the backend supplies the steps, copy and schema version, so web, iOS and Android render one flow. It turned cancellation into a retention flow that records structured reasons for leaving.
Designed and grew a global event-driven service-oriented architecture, moving services from legacy polling to publish/subscribe messaging on AWS SQS and SNS. I also moved us off an early RabbitMQ setup onto AWS-native services, which cut the operational overhead of running our own broker.
The first articles are being written: how this site works, an embroidery renderer, and running LLMs in production without the hand-waving.