DataFramed

#361 The New Data Team with Veronika Durgin, VP of Data at Saks

49 min

In this episode of DataFramed, VP of Data at Saks Veronika Durgin discusses the evolving role of data teams in the age of AI, arguing that while generative AI will automate many technical tasks (like BI report creation and basic data engineering), the human skills of asking the right questions, understanding business nuances, and aligning definitions will become more critical. She champions the rise of analytic engineering as a key future role, emphasizes the importance of conceptual data modeling in providing context for AI, and shares her philosophy on building agile, proactive data teams that are closely integrated with the business.

Summarized by Podsumo

🎧 Listen 🎙️ Ask about this episode

✨ Key Takeaways

1

Veronika predicts that the action of creating a BI report will disappear as tools become easier to use, but that dashboards and scheduled reports will still be valuable for avoiding repetitive queries.
2

She believes the future belongs to 'analytic engineering'—roles that bridge business understanding and technical implementation, focusing on translating business needs into accurate data models.
3

Conceptual data modeling (drawing boxes and lines to represent business concepts and relationships) is highlighted as an irreplaceable human skill that provides essential context for AI and LLMs, reducing hallucination and reconciliation issues.
4

Veronika argues that data teams must shift from a reactive 'help desk' model to a proactive 'customer success' approach, relocating closer to business teams to truly understand their day-to-day challenges.
5

She emphasizes the need for data professionals to stay current in a fast-changing field, suggesting that during interviews she looks for candidates who can explain how they learn outside of immediate work tasks.

💬 Notable Quotes

"A surgeon still needs to know where your heart is. A carpenter still needs to know what the load bearing wall is, even though a carpenter has a drill. They have automated tools, but they don't always necessarily use them."
"If you don't stay current with what's happening, how will you possibly solve new problems in new ways?"
"Data needs to be centralized, but analytics needs to be decentralized. I struggle to see how decentralizing data will work because you have to have tight alignment on definitions and how everybody talks about things."