How can we use AI (responsibly) in Data Science? A tutorial with insights from hands-on practice

Andrea Cappozzo (Professor of Statistics)

Department of Statistical Sciences, Università Cattolica del Sacro Cuore

Abstract

Large Language Models (LLMs) are rapidly transforming how statisticians and data scientists work, yet many remain unfamiliar with their practical integration into data analysis workflows. Drawing directly on my own hands-on experience, this workshop provides an applied introduction to AI-powered tools designed for data science in R and Python.

We will explore three key approaches: 

1. leveraging an AI-integrated IDE for assisted coding and debugging,
2. programmatically interacting with LLMs through R and Python for tasks such as structured data extraction, 
3. using AI to support exploratory data analysis.

Through live demonstrations, we will examine both the impressive capabilities and important limitations of these tools, emphasising their performance while underscoring the critical need for human expertise and oversight.