What happens when you lock a group of AI experts in a room, feed them pizza and give them access to a brand-new AI platform? You get our first Route33 hackathon. And yes, things got a little out of hand.
Over the past few months, our teams have been working hard on Route33, a platform designed to bring clarity to the growing complexity of AI. Through a single access point, Route33 connects organizations to multiple AI models and providers while helping them keep control of costs, governance, security and compliance. The goal is simple: make AI easier to adopt, manage and scale. Before launching Route33 more broadly, we decided to do what any responsible group of AIGENEERS would do.
Try to break it.
So we brought together a room full of AI experts, developers and builders for a Route33 hackathon. Armed with pizza, caffeine and an unhealthy amount of curiosity, they spent the day putting the platform through its paces. Agentic AI, agentic programming, real-world business cases, dozens of models and providers. If it could run through Route33, somebody tried it.
After a few intense hours, the scoreboard looked impressive: more than 500 million tokens consumed across 11,000+ requests. Apparently that's what happens when you unleash a room full of AIGENEERS on a new platform.
The good news? Route33 held up remarkably well.
And that's important, because Route33 was built from a very real challenge facing organizations today: how do you keep visibility, control and flexibility in an AI landscape that evolves faster every month? Through a single routing layer, organizations can work with multiple AI providers while maintaining oversight on usage, budgets, governance and model selection.
The pilot phase is now nearing completion. The lessons learned during the hackathon and the ongoing pilot programme are helping us fine-tune the platform before the next phase.
Route33 is almost ready. More soon. 🚀
%20copy.webp)
