Applied ML at Scale Strategies & Insights
summary
Target describes an Applied ML framework that converts millions of model outputs into safe item data changes: over 20 production models handling a catalog with ~3,000 attributes, and 4–5 million inferences per year pre-GenAI. A model orchestrator builds dependency workflows, batches or parallelizes calls (latency from hundreds of ms to >10s) and caches inputs; a decisions manager applies confidence thresholds so titles auto-update while highlights are flagged for vendor review. Only minimal source-system changes are required.
AI-written summary · about
chapters (5)
- 0:00Introduction and context
- 2:27What is item data
- 9:22ML models for item data
- 14:00Applied ML framework
- 25:04Gen AI impact and summary
related talks
28mAccelerating AI at Scale with Data
27mA Look Inside Applied Machine Learning
37mData Mastering at Scale with AI
24mSmart approach in development and deployment process for ML models
20mIndustrial-Scale Machine Learning
21mHow The Application Of Machine Learning Is Changing The Way Companies Use Data
27mEnterprise AI at Scale: Success Stories and Insights
29mData strategies to enable ML and analytics at scale