Building Talentbookr for a friend

A friend came to me with an idea. Booking live entertainment, a band or a DJ or a string quartet for an event, is weirdly stuck in the past: endless emails, no clear pricing, no easy way to find who's available near you. He could see the product clearly. He needed someone who could actually build it.

That's a role I like. Talentbookr is his idea and his industry knowledge, and my job is to turn it into something real.

The first thing we did wasn't code. It was scope. We sat down and worked out what the first version actually needed to be, which is mostly an exercise in what to leave out. A marketplace can sprawl forever if you let it, so we cut it back to the one thing that matters: put in the postcode of your event and find local talent you can actually book. Everything else could wait.

Then I built it. Phoenix LiveView for the real-time search, Cloudinary handling the photos and video every act needs to show off, and a data model that could grow into the messier features later without painting us into a corner. It's in early access now, which is the honest label for "working, real people using it, still rough in places."

Working this closely with someone on their own idea is a specific skill, and it isn't really a technical one. It's translation. You have to hear what someone means, not just what they say, push back when an idea will cost ten times what it's worth, and protect the thing that made them excited in the first place. Do that well and you're not a contractor taking orders. You're a partner they trust with the whole thing.

I've built enough of these now to know the pattern. Someone has a clear vision and no way to ship it. Bridging that gap, from a sketch on a call to a product people use, is one of the most satisfying things I do.