This unit covers cloud ML services and operational workflows on AWS, focusing on prompt engineering, model deployment/inference, automation of MLOps, and safety/clarity tools for responsible AI. Students will combine multi-modal modeling concepts with practical AWS tools (SageMaker, Bedrock, Clarify) to prepare models for production use and evaluate their trustworthiness.
Learning Objectives
- Analyze and compare AWS ML services (Amazon SageMaker, Amazon Bedrock, and related tooling) to select appropriate solutions for specified use cases and data types
- Apply prompt engineering techniques and design Bedrock agent workflows to construct effective, multi-modal prompts and agent-driven interactions for target applications
- Demonstrate end-to-end deployment and inference workflows by deploying models to Amazon SageMaker and Bedrock, making predictions, and automating repeatable MLOps tasks
- Evaluate model fairness, bias, explainability, and safety by using Amazon SageMaker Clarify and Bedrock guardrails, and recommend mitigation strategies for production readiness
Curriculum
- 1 Section
- 14 Lessons
- 30 hours
Expand all sectionsCollapse all sections
- Unit 3 — ML Services, Inference, and Safety14
- 1.1Choosing the Right ML Techniques for Different Use Cases20 mins
- 1.2AWS ML Services20 mins
- 1.3Prompt Engineering20 mins
- 1.4Amazon Bedrock Agents20 mins
- 1.5Automating MLOps20 mins
- 1.6Amazon SageMaker Clarify20 mins
- 1.7Understanding Data Types in AI Models20 mins
- 1.8Beyond Text: Multi-modal Models20 mins
- 1.9Summary20 mins
- 1.10Summary20 mins
- 1.11SageMaker Inference20 mins
- 1.12Amazon Bedrock Guardrails20 mins
- 1.13Making Predictions Using Trained Models20 mins
- 1.14Prompt Engineering20 mins
Instructor
Leave a Reply