Accelerating AI & ML at scale with Teradata Vantage & SageMaker
summary
Teradata Vantage integrates with Amazon SageMaker to keep feature engineering and large-scale data prep inside a massively parallel database while training and inference run on SageMaker tools. The stack uses SQL pushdown via a Teradata ML Python library, writes to S3, serializes models as PMML/ONNX/MOJO, and supports Clarify, Canvas, serverless and asynchronous inference. Benchmarks claim 4 hours to 10 minutes and 48 hours to 18 minutes speedups; a deployed use case runs 40,000 models to score 7 million welds per day with sub‑10‑minute operator feedback.
AI-written summary · about
chapters (28)
- 0:00Intro
- 0:42Data analytics is a top investment priority for 2022
- 1:33The challenges with analytics and AI
- 2:18Challenges with ML data
- 2:46Barriers to scaling Al initiatives
- 8:07Tens of thousands of customers use Amazon SageMaker
- 8:44Introducing Teradata Vantage on AWS
- 10:54Native object store (NOS)
- 12:29Bottlenecks occur at every stage
- 15:46Democratize features for increased productivity and quality
- 17:18Vantage data integration with Amazon S3
- 20:56Prepare your analytic dataset at scale
- 21:49Customer benchmarks: In-database processing
- 22:46Sampling of Vantage analytics in-database functions
- 23:18Significant language support for Python
- 24:47SageMaker supports every stage of the ML Lifecycle
- 30:24Empower more to build
- 32:55Model governance
- 37:39Integration flow
- 40:26Vantage API integration with AWS analytic services
- 42:26Use case: Personalized care and interactions with a 360° view
- 45:03Predictive maintenance and condition-based maintenance
- 48:26Use cases Engine failure prediction and monitoring
- 49:22Predictive quality and yield Al and ML can help businesses get to the root cause of process-based losses related to quality, yield, energy efficiency, etc. -protecting their bottom line and improving competitiveness
- 50:55Use case: Spot welding inspection on the industrial cloud
- 51:32The welding process
- 52:33Welding robot data analysis to detect abnormal welding
- 53:01Benefits of ML for manufacturing The potential benefits of M. within the sector are huge, and a trusted technology models often partner with experienced vendors to bring onboard smart development teams with data science expertise and corresponding domain knowledge
related talks
40mHow Teradata uses AWS to power enterprise analytics
11mTeradata AI Unlimited: Build the Future of Trusted AI
39mTrain ML models at scale with Amazon SageMaker Training
25mAccelerate NLP training with Amazon SageMaker
32mBuild quality ML models easily & quickly with Amazon SageMaker Autopilot
49mUse Amazon SageMaker to build high-quality ML models faster
19mUsing Amazon SageMaker to develop and deploy ML models
31mTrain and tune ML models to the highest accuracy using Amazon SageMaker