Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore how to build sustainable and cost-efficient generative AI solutions on AWS through agentic workflows in this 54-minute conference talk from AWS re:Invent 2025. Learn to integrate agentic AI systems using Amazon Bedrock AgentCore with contextual memory, asynchronous execution, and on-demand tool invocation to minimize compute waste and maximize efficiency. Discover how Model Control Protocol (MCP) enables secure connections between AI agents, AWS services, and custom tools for streamlined operations. Master optimization techniques including AWS's Trainium and Inferentia2 silicon that deliver 50% better performance per watt, Amazon SageMaker for scalable development, quantization, and speculative decoding. Understand how to implement auto-scaling, batch processing, and spot instances to prevent over-provisioning while maintaining performance. Gain insights into monitoring and cost management using CloudWatch and Cost Explorer to deliver high-performance, low-carbon generative AI solutions that balance sustainability with operational efficiency.
Syllabus
AWS re:Invent 2025 - Sustainable and cost-efficient generative AI with agentic workflows (AIM333)
Taught by
AWS Events