People ask me about the AI a lot. How I used it, what it actually did, whether it really "built" the platform. It's a fair question. But it's not the one I find myself thinking about.
The question I keep coming back to is what actually went into Gigster. Not the tool. The twenty years.
I've spent two decades on every side of this industry - on stage as a performer, behind the scenes running the businesses that make events happen, and in the inbox, chasing the admin that holds it all together. That teaches you things you don't realise you know until you need them. You learn what it looks like when a deposit goes quiet for weeks before an event. You learn what it looks like when an artist arrives at a booking nobody properly confirmed. You learn how much of this industry still runs on WhatsApp threads, email chains and good faith - and how often good faith isn't enough.
But you also learn the other side of it. You learn how many genuinely excellent suppliers never get discovered, simply because there was never a proper place to find them. You learn how many organisers default to the same three people they already know - not because nobody better exists, but because there was no reliable way to vet anyone new. An industry this talented deserved better infrastructure than it had. That was fine, until it wasn't.
I tried to fix it once already. For the first ten years, Gigster ran the way most of this industry still does - phone calls, spreadsheets, and me holding every relationship together myself. In 2016, I hired a software development company to build something better. I described the platform as clearly as I knew how, and handed it over - the way you're told to when you're not the technical one.
What came back wasn't wrong, exactly. It also wasn't what I'd actually designed. Every feature came with a quote. Every change meant another invoice, another few weeks, another version a little smaller than the one I'd described. I ended up with a platform that did a fraction of what I knew the industry needed, because the gap between my vision and what I could afford to build kept eating the vision itself. For years, I assumed that gap was simply the cost of not being technical. I didn't yet know it was possible to close it.
Everything else - every lesson I hadn't yet found a way to build - I filed away as instinct. I didn't think of it as data. I certainly didn't think of it as something I could implement myself.
Then AI changed the arithmetic, and I sat down to try again - properly, this time, without a dev company translating my brief into someone else's code and losing something at every handoff. I was building it myself. And every one of those instincts turned out to be a rule the platform needed - and the rules got very specific, because the instincts were.
The deposit gone quiet became protection: money held safely and chased automatically - politely, and on a schedule. But twenty years on stage also taught me when not to send the reminder: when the client belongs to the artist, not the platform, no email goes out without the artist's say-so. That single rule - knowing whose relationship it is - isn't something you learn from software. You learn it from being the artist.
The act that took a deposit and vanished became a trust ladder: a brand-new supplier's deposit is held until just before the event; a supplier with ten completed bookings is paid within a day. Trust on Gigster isn't claimed. It's earned, one completed gig at a time - exactly the way it's earned in the real industry.
The unconfirmed booking became structure: a quote only becomes real when both sides accept it, and the moment they do, a signed agreement exists. There is no version of the story where two people remember it differently.
The pricing confusion I'd watched play out a hundred times - an artist quotes one number in a chat, the client receives another on paper - became a simple promise: before anything is sent, the artist sees exactly what the client will see. No surprises on either side of the quote.
The brilliant supplier nobody could find became visibility: real categories, real search, pages for the things people actually look for. And the organiser with no way to vet anyone became a standard: every supplier reviewed before they're listed, and every review tied to a real, paid booking - because I've seen what testimonial pages are worth.
Even WhatsApp made it in - not as an enemy, but as it actually is. This industry lives in chat, and it always will. So the platform doesn't ask anyone to leave; it just gives them something better to send. A quote link instead of a promise. A payment link instead of a bank detail typed into a thread.
None of this is complicated to describe. It is genuinely intricate to build, because an industry's worth of edge cases doesn't fit neatly into a form. But that was never really the point. I wasn't designing features. I was giving twenty years of instinct the structure it always deserved.
What surprised me most wasn't the technology. It was realising how much of what felt like scattered experience was, all along, a blueprint I'd been building without knowing it.
AI gave me the means to build it - and it compressed years of work into months. It didn't give me the twenty years, and it couldn't have. This isn't the platform a software company built for me in 2016. It's the one I actually designed, because this time nothing got lost between what I knew and what got made. I don't think what I built should be measured by how quickly it came together. It should be measured by how long it took to know what to build.
The platform is live at thegigster.com. If any of this sounds familiar - as a supplier tired of being overlooked, or an organiser tired of chasing quotes across five inboxes - I built it for you.
Bring Carina to your stage.