The hackathon compressed the complete data workflow into one weekend: understand the material, choose a defensible question, divide the work and communicate the result.
The Department of Network and Data Science at Central European University organised the second edition of its online student hackathon. Participants joined from different countries to work on socially relevant problems through data analysis and machine learning.
Choosing the story
Teams could build a classification pipeline or create an infographic from a supplied dataset. My two teammates and I chose the visual storytelling challenge. We needed to explore the data, identify a meaningful pattern and present it in a form that a broader audience could understand.
The open-ended format was both useful and difficult. A dataset can support many possible narratives, but not every visible relationship is important or reliable. We repeatedly returned to a basic principle: correlation does not establish causation.
Working under a clock
We began by profiling the dataset, checking missing values and comparing possible angles. Once we agreed on the core message, we divided the work between analysis, visual design and supporting research. Regular check-ins kept those parts aligned.
Under time pressure, agree on the question and the definition of done before polishing the output. A team can move quickly in three different directions and still make no progress.
What the result taught us
After 48 hours of analysis, debugging and design, we submitted a completed infographic. The final artefact mattered, but the process revealed more durable lessons. Good teamwork depends on explicit ownership. Good visualisation depends on hierarchy. Responsible analysis depends on stating what the data cannot prove.
The international setting also changed the experience. Different perspectives challenged assumptions that might have gone unnoticed in a more familiar group. That made the collaboration as educational as the technical work.
I left the hackathon more confident in my data skills, but also more cautious about interpretation. Data science creates value when technical analysis and clear communication reinforce each other. The weekend made that principle tangible.