In a recent episode of the Datacast podcast, Eugene Yan sat down with host James Le to talk about what makes data science work in practice. The conversation spans Yan's hands-on experience, from leading teams to keeping machine learning models running reliably in production. He also touches on how agile methodologies apply—and sometimes clash—with the exploratory nature of data science.
Yan, known for his writing on applied ML and engineering culture, shares lessons on balancing technical depth with leadership responsibilities. The episode also explores how clear communication and documentation help bridge the gap between research and deployment. While the summary is brief, it points to a recurring theme in Yan's work: effective data science is as much about people and process as it is about algorithms.
Listeners looking for actionable advice on production ML, team management, or building a public writing practice will find relevant takeaways. The episode does not dive into specific case studies or metrics, but it offers a high-level map of the skills and habits that sustain long-term success in the field.