It started with one question
When generative AI became mainstream, every industry started asking the same question.
How should we use AI?
The betting industry was no different. Around the same time, several local products had already introduced AI-powered chat experiences, while global companies were rapidly experimenting with new ways of integrating AI into their products.
The pressure to "have AI" was growing everywhere.
For us, however, the question wasn't simply whether we should build a chatbot.
It was whether we could build something that would genuinely improve the experience for sports fans instead of becoming another feature people would try once and forget.
What if?
One day my manager invited me to join a brand-new initiative.
There was no roadmap, no MVP, No validated solution - only a single idea - let's build an AI assistant for our sportsbook.
That uncertainty was exactly what made the project exciting. At first, it felt like almost anything was possible. Personalized conversations, instant search, context-aware recommendations, smarter navigation. Features that could completely change how users interacted with sports content. For a moment, the possibilities felt almost endless.
Instead of immediately discussing APIs, technical limitations, or legal restrictions, we intentionally ignored them.
Not because they didn't matter, because it was too early.
Before deciding what was possible, we wanted to understand what would actually be valuable. So we asked ourselves a much simpler question: If there were no limitations, what would the ideal AI assistant look like?
Over the following workshops, everyone on the team contributed ideas. Design Engineering Product Business Sports experts
Each person imagined the assistant from a different perspective. By the end of this phase, we had generated around fifty different use cases.
Some were ambitious, Some were surprisingly simple, Some would never make it into the product.
And that was completely intentional.
