People say on-demand routing is complicated. We know how right they are.
Managing on-demand fleets involves juggling a lot of moving parts. That’s true. But to anyone who’s tried to do it, it’s also a laughable understatement.
Every moving part behaves unpredictably.
Changing conditions make estimates meaningless.
Closest isn’t always fastest, and fastest isn’t always optimal.
You’re not just solving for multiple variables at once.
You’re solving for how each affects the others, too.
Getting on-demand routing right, at scale, means adapting constantly.
That kind of real-time, coordinated reactivity requires one heck of an algorithm.
We ought to know; we built it.
We started from the ground up with a founding team of rideshare veterans, transit experts, and the algorithm engineer who literally wrote the book on the topic.
5 years and more than 3 million rides later, we’re proud to say our road-tested tech powers the most efficient on-demand fleets on the planet.
And we have miles and miles to go.
