Interview with Devashish Khulbe: Mentoring Datathon at Masaryk University’s Digital Talent Lab

Devashish, from your perspective, what is the vision behind the Digital Talent Lab, and what are the main goals for participants?
The Digital Talent Lab grew out of our Digital City Lab research group. Our goal was to invite more people to collaborate on ongoing projects or even bring their own ideas. For participants, the main objective is to gain hands-on experience in applied data science and AI, especially as it relates to real-world urban challenges. We want them to test and develop their skills while contributing to meaningful research.

What sets the Digital Talent Lab Datathon apart from other data science or AI training initiatives?

Our Datathon offers a truly hands-on experience with real-world projects and datasets sourced directly from the Digital City Lab. Participants spend three weeks working on easy or medium-level projects, and five weeks on more challenging ones. Unlike typical online hackathons—which often last only a few days and require you to work alone—our mentors are actively involved throughout the process, providing guidance, answering questions, and offering regular feedback. We hold frequent update calls and are always available by email, ensuring a supportive and collaborative environment. Best of all, participation is completely free, making it an excellent opportunity for anyone eager to launch their career in data analytics, ML and AI.

Can you tell us more about the types of projects participants work on?

Most projects focus on applied data science and AI for digital and smart cities. We use real urban datasets—like healthcare movement data from the Czech Republic or geolocated Twitter data from New York City—to solve practical research questions. For example, participants might analyze how people move for healthcare or how social media sentiment varies across urban neighborhoods. Many projects are direct spin-offs from our ongoing research at the Digital City Lab, so participants work with the same data and tools as our researchers.

What skills are most important for success in the Datathon?

A background in mathematics is helpful, but you don’t need to be an expert coder. With today’s AI tools, even beginners can learn to code and analyze data. High school-level math and some programming basics—like Python—are enough to get started. What matters most is curiosity, a willingness to learn, and the ability to ask good questions.

How do you approach mentorship during the Datathon?

Mentorship is central to our approach. We stay in constant touch with participants, holding weekly meetings to discuss challenges and progress. We review code, provide feedback, and help them overcome obstacles. In previous editions, we even guided teams through collaborative phases, and the feedback has been overwhelmingly positive.

What impact do you hope the Digital Talent Lab and Datathon will have on the broader AI and data science community?

Our aim is to train and inspire people to enter the fast-growing AI field. By working on real projects, participants gain practical skills and experience that can lead to internships, research positions, or even PhD opportunities in the Digital City Lab. Ultimately, we hope to attract talented individuals who want to contribute to urban data science and AI research.

What is the main research focus of the Digital City Lab, and what is your role?

The Digital City Lab originated at the New York University as the Urban Complexity Lab led by prof. Stan Sobolevski, focusing on urban research questions using big data, machine learning, and AI. After moving to Masaryk University, the mission remains the same: to analyze and solve urban problems with advanced data science methods. I joined the lab as a master’s student and continued through my PhD, working on projects involving urban mobility, healthcare data, and social media analytics.

Can you share an example of research from the lab that has had real-world impact?

One project analyzed the effects of congestion charging in New York City. We found that such policies disproportionately affect low-income residents, leading to significant shifts in transportation choices. Insights like these are valuable for policymakers considering similar interventions in other cities.

What motivates you personally in this work?

I’m passionate about analyzing large datasets to answer real-world questions, especially in urban contexts. Recently, I’ve focused on graph-based machine learning, which is powerful for modeling complex urban systems. The field is always evolving, so continuous learning is essential—and that’s something I really enjoy.

How is your collaboration with Stan, the head of the lab and your PhD supervisor?

I’ve worked with Stan for over five years, both at NYU and here at Masaryk University. He brings a unique perspective from mathematics, data science, and urban science, which is incredibly valuable for our interdisciplinary research. His mentorship has been instrumental in my development as a researcher.

Thanks for the interview.