Python and the AI Lakehouse
summary
Claims that Python becomes native to the Lakehouse by adding a feature store with online sync to operational DBs, a vector index and model registry, and an AI query engine. The team builds a Hop Query service using Arrow Flight zero-copy to read/write tables from Python and reports 9 to 45x throughput versus other cloud clients. Feature freshness, low‑latency retrieval, monitoring and MLOps pipelines are implemented and shown to work end to end.
AI-written summary · about
chapters (8)
- 0:00Python and AI Lakehouse intro
- 3:50Traditional lakehouse limits
- 10:38Extending to AI lakehouse
- 12:34Python API for features/models
- 16:50AI query engine for Python
- 20:29Python AI pipelines
- 21:55AI lakehouse benefits
- 23:06Q&A
related talks
28mBridging Python and Lakehouse worlds with Arrow Flight and DuckDB.
31mOpen Standards and Lakehouse AI Development
01h 10mReal-time AI Lakehouse
23mBuilding AI-Driven Real-Time Lakehouse APIs
34mLeveraging Apache Arrow for ML Workflows
50mOleksandra Bovkun | Data Track | Compass AI & Data Summit 2025
39mPython and SQL: Better Together, Powered by DuckDB
29mLakeside AI: Bringing Models to Data with Open Lakehouses & Agents)