Michael Hardegger
Team Leader
Michael is a data science professional with 15 years of experience in applying AI to business and research problems. He leads a cross-functional data team at Digitec Galaxus AG that focuses on data-driven automation of purchasing, pricing, and related decisions.
From Illusion to Impact: A Journey through Data Science Missteps
It starts with a business stakeholder and a vague belief that data science can solve a poorly defined problem. Add to this mix a project manager keen to impress and an AI engineer with a PhD in graph neural networks but little interest in the business context.
This sets the stage for what we call the “data science cycle of garbage”. This cycle produces fancy graphics, extensive reports on model selection (spoiler alert: you will likely end up with a graph neural network, regardless of the actual problem), and managers proudly labeling their company as “data-driven”.
In this talk we’ll discuss why this cycle is so common and how to mitigate its impact on a company’s journey towards effective AI implementation.