Build Smarter, Don't Just Build
summary
A four-person product data science team scales insights across dozens of clouds by automating three bottlenecks: cataloging raw logs with Elation to cut data-prep from weeks to hours (new hire goes from zero to running analyses in days), automated feature discovery with orgDNA that finds subpopulations via F-test–style contrasts, and reproducible experimentation with Domino plus a homegrown model-monitoring service that alerts on drift so models stay useful in production. Numbers cited: 30→270 orgs, weeks→hours, team of 4 delivering production models.
AI-written summary · about
chapters (6)
- 0:00Data science at Salesforce
- 6:48Productivity challenges
- 9:58Data preparation tools
- 15:09Automating feature selection
- 16:41Continuous experimentation
- 19:56Model monitoring and closing
related talks
31mProductionalizing a Data Science Team
38mData Secrets from a Platform Engineer
35mBuilding Data-Intensive Teams
18mHow to create a high performance data team: lessons learned from the field
32mTeam Teardown: Evolving an ML Platform
14mBuilding a Data-Driven Culture: An Anthology of Strategies and Insights
26mData, It’s a People Problem: Understanding the Sociotechnical Elements of a Data-Driven Organization
32mMaking Models Work: How an optimized ML lifecycle can drive better ROI