


Public transport app designed to plan journeys using real time data, scientific algorithms and machine learning

Our combination of scientific algorithms, real time processing of traffic situations and crowd-sourced reports allows us to predict journey durations and arrival times more accurately than our competitors and local transit authorities. Data quality and accuracy is the critical component in this business and we’re not compromising on it whatsoever
A self-learning algorithm helps us to understand the local commuting experience. Machine learning works by understanding local patterns and provides unique results and preferences for different cities and different users without the need for manual data tuning

The beauty here is not only that we allow users to create any public transport related content, but also that our search algorithm filters and takes into account what’s important to them, so we can notify other fellow commuters and dynamically adjust routing. Crowd-sourcing at its best!

Offline mode enables us to make transit better without an internet connection. Because we’re no longer dependent on a data connection, the app is faster than ever and takes barely any space. Everybody who lives with poor or expensive data connections can rely on the most accessible TRAFI’s advice.


Martynas, the first of his name, is the King of TRAFI and protector of the realm. ‘GoT’ fan obviously, snowboarder and aviation geek. Believes he can land an A380 safely...

Few keywords to describe Jurgis: International computer science competitions, MIT, theoretical physics, PhD, CTO, string theory, squash, Belgium beer, enough?

Bunch of crazy and brilliant minds from all around the globe devoted to improving public transport worldwide! Wanna make an impact?





