PyData Berlin: How The Data Community Comes Together - Feb Edition

The February edition of PyData meetup was held at GetYourGuide's Ampere Building, our Berlin office, and it was an enlightening event for data enthusiasts to share their experiences and findings. PyData is a non-profit community that organizes educational meetups across various locations to discuss all things data and promote mutual learning opportunities. Our Senior Data Manager, Theodore Meynard, hosted the meetup, starting with a brief explanation of the agenda and the rules of conduct.

The first talk of the day was presented by Andrey Cheptsov, the creator of dstack, an open-source tool that facilitates ML experiments. He spoke about "Bringing GitOps to ML experiments with dstack" and demonstrated how GitOps, a DevOps practice, could improve the software shipping process. Andrey's commitment to his work was evident in his ability to break down complex data topics quickly and easily.

Next up were three data scientists, Anand Deshpande, Jan Meyer, and Onur Kerimoglu, who presented on "Using Graph Neural Networks to predict traffic accidents."They discussed the frequency of traffic accidents happening in Berlin, evaluated past data in terms of area and population, and represented it visually using a graph model. The idea was to combine neural networks with a graph model to recognize and understand non-linear relationships.

{{slider}}

After each presentation, attendees had a chance to ask questions, and we encouraged them to network with the speakers directly to deepen their learnings. The PyData group organizes monthly meetups, and it's an excellent opportunity to get a glimpse into the work of fellow data practitioners. Overall, the PyData meetup was an excellent forum for data enthusiasts to come together, learn, and network. We look forward to seeing you at the next one!