01 · Where I am coming from

Your roadmap is the bigger version of something I have already built

I spent two years at Quad Lock looking after the video and photo library every team relied on, from social through to the boardroom. Your position description honestly reads like the next size up of that work, and I would really love to help you get there.

• Looked after a library of more than 500,000 files, and labelled 80,000 of them by hand before automating anything. Doing it manually first is how I learned which labels actually mattered
• Built the AI that took over the labelling, with a person checking its work before anything was trusted. It was about three times faster, and nothing went live until it was accurate enough
• Built a Slack bot so eleven teams could simply ask for footage in plain English instead of hunting through folders. That is what made people actually use it
• Wrote the procedures, guides and short screen recordings so the whole thing kept running without me

02 · What I would do here

Your five areas, in plain terms

Your position description already sets out a roadmap. I would deliver that rather than replace it, and the first thing I would ask for is a walk through it.

1. Make things findable. Look at what is actually in the library, fix how it is organised, and agree how footage gets labelled from here on
2. Ask people what they really need. Watch how they look for footage rather than survey them, because what people do and what they say they do are different things
3. Smooth the handovers. From shoot to edit to approved, with fewer manual steps and less work being redone
4. Cut the waste. Review the tools and subscriptions, find what overlaps, and automate only the parts that can be checked
5. Write it down properly. Procedures, short guides people actually open, and clear rules on what gets kept and for how long

03 · How we would know it worked

Five things worth counting

Measured at the start so any improvement can be shown rather than claimed.

• How long it takes to find something. The simplest number, and the one nobody has
• How often footage gets shot again because nobody could find what already existed. This is the most expensive problem and it never shows up on a software bill
• How many “where do I find…” messages the team gets in a week
• What the tools cost against what is actually used. Unused seats and overlapping subscriptions are the quickest saving available
• Whether anything has quietly expired. Music licences, talent releases and permits should be flagged before they lapse, not after