Designing the flow
The acquisition pipeline I took on was functional but built around assumptions that were starting to limit us — local dependencies, siloed per-show storage, a stateless process with no persistent record of what moved or why. The redesign was less about replacing it and more about rethinking the principles underneath it: reduce ground dependencies, open the architecture for cleaner interfacing with downstream systems, and build an archive structure intuitive enough that assets land where they belong automatically — no manual transfers, no tape ceremonies, no one deciding where something goes. The result was a unified S3 schema spanning a decade of archive and live delivery, a system of record in DynamoDB that made every asset and transaction traceable, and automated processing that delivered edit media, proxies, and DAM-ready files simultaneously — each fingerprinted with show, episode, format, and a unique ID. Those early design choices are what made every subsequent phase of the modernization possible.
- Cloud-first acquisition portal on Aspera with REST API automation
- Unified S3 path schema spanning a decade of archive and live delivery
- DynamoDB system of record; SQS as the cross-system message bus
- Orchestration via CatDV, Vantage, and Marquis MEWS/Medway
- OpenText DAM redesign — storage-in-place, title-driven navigation, 1.8M assets
- Jira-driven media governance — flight dates, licensing windows, revision states