Ravi Bapna, da Carlson School of Management, University of Minnesota, vai apresentar o seu trabalho de investigação. 

Agentic AI and Managers’ Analytics Capabilities - An Exploration

As companies generate increasing volumes of data, their ability to extract actionable insights remains limited, with estimates suggesting that only a small fraction of available data is effectively utilized. While data scientists play a crucial role in business analytics, their impact is constrained if their managers lack the ability or inclination to engage with analytics tools. Many managers (often MBA graduates), particularly those without strong quantitative backgrounds, struggle with coding in R or Python, limiting their ability to gain proficiency in the full capabilities of modern business analytics. This study examines whether a Gen-AI-powered virtual data science assistant, DatBot, can enhance MBA students’ analytics proficiency. Using a Retrieval-Augmented Generation (RAG) architecture, we integrate DatBot into Slack teams for 400 students enrolled in a core analytics course at a top 20 global business school. Our primary outcome variable is students’ final exam scores, which serve as an important academic and career milestone. Our findings indicate that greater engagement with DatBot leads to a significant improvement in exam performance, with a doubling of prompt volume increasing scores by 1.5 percentage points—a meaningful impact given the role of core course CGPA in securing interviews with top consulting firms. Additionally, we find that low performers benefit the most, suggesting a leveling-up effect rather than the reinforcement of existing disparities (Matthew Effect). These results highlight the potential of agentic AI to bridge skill gaps in business analytics education, better equipping managers with capabilities to make data-driven decisions.