ML feature store a cornerstone in modern data architecture for AI/ML
summary
Argues that Redis and Redis Enterprise serve as the low‑latency backbone for feature stores, describing four concrete manifestations: online feature store for sub‑millisecond serving, offline store for training, model store for caching binaries, and an evaluation store for monitoring and retraining. Cites a Forrester joint study: ~40% of decision makers expect current architectures to fail future needs and 64% report 20–39% of models use real‑time streams. Gives production examples (DoorDash uses Redis for prediction features; AT&T pairs MLflow, Delta Lake, MongoDB/Hive, and Redis) and highlights Redis on Flash for larger datasets.
AI-written summary · about
chapters (7)
- 0:00Introduction to ML feature stores
- 3:57Redis technology overview
- 8:12Real-time requirements for AI
- 10:18Evolution of ML architectures
- 12:02Redis as feature store
- 14:09Customer use cases
- 17:42Redis for real-time ML
related talks
19mThe Key Role Of Feature Stores In Deploying Real-time AI
43mThe Challenges of Deploying Real-time AI for Finance and how Open Source can help
08mFeature Store Usage Patterns - From a Single Data Scientist to an Enterprise
44mFeature store
14m5 great ways to reimagine your data layer
35mPanel 03- How to make the Feature Store become a hidden part of the productionalization of ML models
34mSolving the ML Pain You Forgot to Mention
40mCreating and operating ML models from event based data using feature stores