High-performance data foundation for AI success
summary
Argues that a data platform must act as a business operating system and shows a concrete stack: Apache Iceberg cataloging a lakehouse, governed by AWS Lake Formation, with Amazon SageMaker and Bedrock for model and agent work. Recommends zero‑ETL/zero‑data‑sharing options, APIs for agent context, and delivering an initial MLP in 60–90 days. Gives a concrete first use case—an agentic customer support flow fed a propensity model and a summarizer of the last five calls—and suggests iterative flywheel scaling.
AI-written summary · about
chapters (11)
- 0:00Introduction and audience setup
- 1:06Evolution of data strategies for AI
- 7:30People process technology alignment
- 10:43Rethinking the data platform
- 13:16Designing for business personas
- 15:43AI agents and emerging personas
- 17:08Core data platform elements
- 20:42Bridging to production
- 23:51Business operating system considerations
- 25:15Unifying data on AWS
- 34:01AWS programs and closing
related talks
47mDataOps for AI-Ready Platforms
43mBuilding an AI-Ready Data Foundation
26mThe Hidden Foundation: How Effective Data Management Determines AI System Success
31mBuilding a Data Infrastructure for AI/ML
30mEnterprise AI: How To Operationalize Data Analytics At Scale
47mBuild cloud data management across analytics, ML & gen AI applications
01h 17mData Engineering in the AI Era: Building a Robust Foundation for AI Applications
26mEngineering Excellence: Platforms and Architectures for Real-Time AI Analytics