Harvard Data Science Review Podcast Tracking the Most Intoxicating Data: A Conversation With Eric LeVine
Nov 20, 2025
Eric LeVine, the founder of CellarTracker and an ex-Microsoft program manager, shares insights into the intersection of wine and data science. He recounts how CellarTracker evolved from a personal project to a global wine database with millions of users. LeVine discusses using AI to predict wine preferences and identifies 'digital twins' for personalized recommendations. He emphasizes the importance of maintaining human elements in wine discovery and privacy principles in data usage. Dive into how technology is transforming the way we experience wine!
AI Snips
Chapters
Transcript
Episode notes
From Tuscany Trip To Global Platform
- Eric LeVine started CellarTracker as a personal tool after a Tuscany trip and spreadsheet frustrations at Microsoft.
- He launched it in 2003 on sabbatical, shared it with friends, and it scaled rapidly into a community platform.
Small Core, Large Research Audience
- CellarTracker's core users are a small, highly engaged collector audience generating rich data.
- Most site traffic, however, comes from people researching wines, not cataloging cellars.
Collectors Represent Big Economic Value
- Collectors recorded on CellarTracker represent a large economic value, often billions over time.
- The platform reveals wide diversity of taste and changing individual journeys in wine preferences.