← back
Ship it! Building Production Ready Agents — Mike Chambers, AWS
Takeaway
Bedrock Agents productionizes the five components of an agent (model, prompt, loop, history, tools) without managing GPU infra.
Summary
- AWS DevRel walks anatomy of an agent: model, prompt, loop, history, tools — each must be cloud-scale to ship a production agent.
- Demo: a Llama-3.1-8B dice-rolling agent ('roll for initiative, add +5 dexterity') runs locally then is migrated to Amazon Bedrock Agents (fully managed, no infra).
- Bedrock provides models from Anthropic, Amazon Nova, Meta, Mistral, AI21 plus Bedrock Agents for the orchestration layer with editable prompt templates.
- Conversational history is framed as critical not just for cross-session recall but for the agent's intra-loop reasoning steps.
- Course context: AWS+DeepLearning.AI 'Fundamentals of LLMs' has 370K+ learners.
bedrockagentsaws
Original description
Explore the practical challenges and solutions for deploying AI agents in real-world production environments. Through detailed technical analysis and practical examples, we'll examine strategies for building and orchestrating agent systems at scale. We'll cover critical infrastructure decisions, scalability frameworks, and best practices for creating robust, production-ready agent architectures. About Mike Chambers Mike is a passionate developer advocate and expert in the fields of machine learning, AI, and generative AI. With a strong background in cloud computing, he brings a unique blend of technical knowledge and communication skills to engage and inspire audiences. Join Mike in exploring the possibilities and shaping the future of these transformative technologies. Recorded at the AI Engineer World's Fair in San Francisco. Stay up to date on our upcoming events and content by joining our newsletter here: https://www.ai.engineer/newsletter