Scaling your data and AI from 0-100 with open sourcePushed backThe speakers disputed the assumption that production feature pipelines should be rewritten and maintained by a separate data or ML team after data scientists hand over notebooks.52:54
Feature stores, and second thoughts in 2025
6 sessions
Scaling your data and AI from 0-100 with open source
Hard Learned Lessons from Over a Decade in AIClaimTecton's feature store centralized and automated the data pipelines needed to create training sets and provide real-time model features.5:07
PodcastReal-time Feature Generation at LyftPushed backA conventional feature store is not sufficient for Lyft's hierarchical, geospatially aggregated features.23:57
Prepare Your Data for AI Now, or You're Screwed LaterClaimThe difference between classical ML features and agent context is that classical features are usually hardcoded into a model, while agents must discover the information they need.18:38
Traditional vs LLM Recommender Systems: Are They Worth It?ClaimTraditional recommender systems require explicitly selected signals and hand-crafted features, while LLMs can infer which signals matter from the user, item, and interaction context.4:52
PodcastContext Engineering 2.0Pushed backFeatureform's acquisition by Redis does not signal the death of the feature store.0:25