Fanatics swapped a £125k-a-year licence for a platform they own.

Fanatics needed to modernise several business-critical systems at once. The most pressing was a third-party warehouse communication platform approaching licence renewal — a system that sat directly between ecommerce sales and fulfilment, cost around £125,000 a year to keep, and could not be switched off while orders were flowing through it.

Renewing meant paying again for a dependency the business didn't control. Replacing it meant rebuilding a live integration under a renewal deadline, with no acceptable window for downtime.

Running in parallel was a second problem: roughly 1.5TB of ecommerce imagery needed to move to AWS S3, and every image reference across client deployments had to follow it. Those deployments were the storefronts of some of the most recognisable organisations in sport — F1, Manchester United, the NFL, the IOC, MLB, AFL, Wimbledon and PSG. A broken image on any of them was a visible failure.

Bringing warehouse communication in-house

I led a team of four developers designing and implementing an in-house TCP/IP communication platform linking the ecommerce sales systems to warehouse fulfilment operations. Rather than treat it as a like-for-like rewrite, we worked with stakeholders to define the delivery plan and migration strategy up front — so the sequencing, not just the code, was agreed before anything moved.

Shadow deployment instead of a cutover

The new platform ran alongside the licensed one, handling the same traffic in shadow while the old system stayed authoritative. That gave us production evidence rather than test-environment confidence, and it turned the switch itself into an anticlimax — which is what a migration on a business-critical path should be.

Migrating 1.5TB and every reference to it

Separately, I architected the backend platform migration supporting the relocation of around 1.5TB of ecommerce imagery to AWS S3. The hard part was never the bytes — it was the references. I designed a RabbitMQ messaging pipeline to update image references across every client deployment, and directed implementation across the team so the work proceeded site by site rather than as one irreversible sweep.

Full-stack development and team leadership in an Agile/Kanban environment, spanning microservices design, multi-site deployments, and both the warehouse-integration and image-migration projects.