The Data Science Education Podcast · Berkeley Data Science

Preparing Data Science Undergrads for Industry (feat. David Chew and Vik Gopal)

March 10, 2023·30 min·2 clips

Eric and Kalechi from UC Berkeley kick off a conversation with Vic and David Gopal about preparing data science undergrads for industry.

This episode of the UC Berkeley Data Science Education Podcast features hosts Eric Van Dusen and Kalechi Nebedam in conversation with Vic Gopal from the National University of Singapore. Vic holds a PhD in statistics and has industry experience from IBM before transitioning to academia. The discussion centers on preparing undergraduate data science students for careers in industry, drawing from Vic's decade of teaching experience. Vic teaches a course nominally on visualization but fundamentally about communicating data insights effectively. He emphasizes that students often struggle to translate technical analyses into compelling business narratives. The program at NUS incorporates significant project-based learning, requiring students to tackle problems with real-world data. A key initiative involves partnering with local companies to provide students with authentic industry challenges. Vic argues that academic programs must bridge the gap between statistical theory and practical application in business contexts. He shares observations from his time at IBM regarding the skills new graduates typically lack. The conversation explores how capstone projects simulate professional data science workflows, from problem definition to presentation. A surprising claim is that mastering visualization tools is less critical than developing the judgment of what story the data tells. Vic suggests the most valuable skill for students is learning to ask the right questions of a business stakeholder. He notes that industry partners often provide messy, unstructured data, which becomes a core part of the learning experience. An interesting insight is that students frequently over-engineer technical solutions before understanding the core business need. The pedagogical approach prioritizes iterative feedback from both instructors and industry mentors. Memorable moments include Vic contrasting the academic pursuit of model accuracy with the industry priority of actionable results. The tone is informal, conversational, and focused on practical pedagogy. The style is interview-based, with hosts asking questions to draw out the guest's experiences and recommendations. This episode would appeal to data science educators, curriculum designers, and students curious about industry expectations. Listeners seeking highly technical deep dives or debates on theoretical frameworks might find it less aligned with their interests.

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from The Data Science Education Podcast