Drive transformation through machine learning
summary
Company migrates analytics to Amazon SageMaker, using 48-hour sprints and shared Jupyter notebooks to iterate models quickly. Pipeline reads from S3, trains in SageMaker, and exposes predictions via Lambda-triggered SageMaker scoring and HTTP endpoints while logging results for accuracy tracking. A production system uses QR-code image captures to score cars, applies linear regression and slope checks to rank cars for redeployment, and measurably reduces idle fleet time.
AI-written summary · about
chapters (12)
- 0:00Intro
- 1:39Today's key takeaways
- 2:17Why AWS & Slalom?
- 4:27Machine learning (ML) in the cloud
- 6:37We need managed ML
- 7:28Amazon Sage Maker enables business transformation
- 9:11Define your "why" with a few questions
- 9:56Drivers behind what models to migrate
- 11:16Options for migrating analytics workloads to AWS
- 15:38Our model development followed an agile, sprint-based process
- 16:34Mileage optimization ML architecture
- 17:36The model takes key operational components into consideration for deployment
related talks
43mMachine Learning Ops: DevOps for Data Science
19mUsing Amazon SageMaker to develop and deploy ML models
32mAccelerating autonomous vehicle solutions with SageMaker, featuring Lyft
37mStreamlining Machine Learning Operations
31mPut Your Machine Learning on Autopilot
30mOperationalizing Machine Learning
32mBuild quality ML models easily & quickly with Amazon SageMaker Autopilot
37mStreamlining Machine Learning Operations