Accelerating AI at Scale with Data
summary
Intuit cuts model development from six months to under a week by rebuilding cloud data and ML platforms: a secure multi-petabyte lake with unified ingestion, paved‑road persistence, real‑time stack (Kafka/Flink/Beam), a pipeline orchestration and curation layer, and a feature engineering platform. Operations run hundreds of production models with 10% of models trained daily, tens of thousands of features and billions of updates, processing hundreds of thousands of events/sec and yielding 50% fewer operational issues.
AI-written summary · about
chapters (7)
- 0:00Introduction to AI at scale
- 3:52Reducing AI development time
- 7:32Modernizing data infrastructure
- 12:36Data ownership challenges
- 17:52Future of data infrastructure
- 20:19Key strategies for acceleration
- 22:04Q&A session
related talks
23mAnalyzing Operational Data at Scale Using ML at Intuit
26mManaging ML Models @ Scale
33mLeaning into the Data Mesh
37mOn Building Production-Quality AI Systems
35mHigh-performance data foundation for AI success
16mOperationalizing AI in Finance: Real-Time Analytics
23mAI-Driven Enterprise Intelligence
34mFast-Track Your AI-ML Model Integration