Everything that has to happen outside the request/response cycle — translation, image moderation, video encoding, thumbnail generation, AI orchestration.
Azure Functions handle the spiky, asynchronous work — translation requests, image moderation, video encoding. Each scales independently on demand and never touches the request/response path.
Translation worker. A request lands in the is-translation-generic-function queue; the function picks it up, calls the translation service, caches the result, and logs the job run.
Auto-scans uploaded images for unsafe content via Azure AI Content Moderator. Flags or removes the media before it ever reaches an end user.
Video transcoding pipeline. Event Grid fires when a video blob is uploaded; the function runs it through Azure Media Services and writes back streaming URLs.
Azure WebJobs handle the steady-state workloads — bulk translation sweeps and continuous thumbnail generation. They run in the background without blocking the live API or burning through translation-vendor quotas.
Bulk and scheduled translation sweeps. Reuses the same translation service the Function uses, but operates on large batches across every supported language.
Continuously generates thumbnails and full-size variants for every image in blob storage. Output dimensions are configurable via app settings — no code changes needed.
A test harness for specific features and projects. Can be wired into the CI/CD pipeline to run on every push so a regression never reaches a deployed environment.
The automated test suite. Tests are added here for specific bugs or regressions we want to guard against. The project can be run locally or wired into the CI/CD pipeline as a pre-deployment gate.
Book a 30-minute architecture review. We’ll trace the queues, the Functions, the WebJobs, and the shared library layer end to end.