Broadcast & media group
Two thousand builds, one pipeline
A media group with teams on different build systems and no shared way to release. Consolidated onto one platform, with the bespoke tooling needed to make the move survivable.
The situation
A national broadcast and media group — television, radio, print and online — with engineering teams that had each chosen their own build system. Moving between teams meant relearning how to ship. Nobody could answer basic questions across the estate: what is deployed where, who released it, and when.
What was done
Migration of the whole organisation from one build platform to another. The interesting part was never the platform — it was that hundreds of existing pipelines had to keep working while it happened.
That meant writing the tooling: translation scripts to convert existing job definitions, verification that compared old and new pipeline output before anything was switched, and shared pipeline templates so teams stopped solving the same problems separately.
Alongside it, DNS and certificate management were automated. Both had been manual, calendar-driven tasks — the kind that work fine until the person who remembers is on holiday. Video ingest infrastructure was rebuilt as code in the same period, which took the deployment of that pipeline from a day to minutes.
Where it ended up
One build platform, shared templates, and a measurable increase in deployment frequency. Certificate expiry stopped being an incident category.
The lesson that generalises: a migration between tools is mostly a migration of habits. Budget for the tooling that lets people move without relearning everything, or the old system never actually dies.