In this workshop, we slip into the role of a data scientist: Faced with an unknown dataset, we learn how our research prototype Dataloom makes it possible for anyone to explore, clean, and analyze data without any programming skills or prior knowledge of the dataset. Workshop participants get hands-on experience with Dataloom, learn how an agentic data pipeline works, and learn interesting insights about the data we are exploring together!

Agentic data agents are becoming more and more popular. Existing tools like Excel, PowerBI, or Tableau are starting to ship AI features. New systems such as WrenAI, Datajar, or Dataloom are built from the ground up as agentic data agents. The idea is to leverage generative AI (that is ChatGPT, Claude, etc.) to make data science accessible to non-expert users. Thus democratizing access to data, which is hugely important in professional as well as private life:

In this workshop, we will use an agentic data agent and see first hand how simple it can be to explore data. We will see its advantages and also discuss potential pitfalls, such as hallucinations. Each participant will get access to a Dataloom instance and gets to explore a publicly available dataset.