When I joined Ownzones in 2015, Paramount was still shipping feature films to distribution partners on physical hard drives via courier. We replaced that with a cloud dashboard.
In 2015, high-resolution video files were large enough, and internet infrastructure unreliable enough, that the dominant way major studios delivered content to partners was still physical: a hard drive, boxed, shipped by overnight courier, signed for on the other end. Every step of that pipeline had a human doing something software could do.
Ownzones was building the cloud-native replacement. I joined as the company’s first UX hire, establishing the design function while designing two products in parallel: Connect (the operator-facing admin platform used by studios and post-houses) and the consumer VOD app.
On Connect, the work was synthesis: translate messy, half-offline studio workflows into a clean interface that made those workflows continuous and visible. The backend was the UX — deep time with the engineering team to understand the transcoding and delivery architecture, and sustained stakeholder conversations to surface what the physical process was actually doing.
1. Design the dashboard around the job, not around the system. The engineering architecture of a cloud transcoder is a tree of jobs, parameters, output profiles, and delivery endpoints. The mental model of a studio operator is simpler: I have a file, I need this partner to receive it in the right format by this deadline. The interface that shipped started from the operator’s job. Job view was default; system view one click away for engineers and power users.
2. Earn the persuasion round trip. Iteration — calls, focus groups, workshops, wireframes, rework — was the baseline I held the team to, and it didn’t come naturally to waterfall-trained studios or to engineering focused on transcoder performance. The persuasion happened one feature at a time; the quality of what shipped was the argument. That discipline — defending iteration to senior, time-pressured stakeholders without losing momentum — is the single habit I carry into every complex B2B project since.
Operator tools that replace physical processes have a specific honesty: you can point at the old process, count its friction, and design against the count. The best operator UX is frequently not in the features you add but in the steps you remove.