Agentic AI Engineering on AWS

About this course

AI agents are everywhere. Production AI systems are not.

Most courses stop at prompts and demos. This course teaches you how to design, build, and deploy a production-grade agentic AI platform on AWS, the same way engineering teams build real systems at scale.

Across 31 hands-on lessons, you will build a complete multi-service platform from the ground up using Python, AWS Bedrock (Claude and Titan models), Terraform, Kubernetes (EKS), FastAPI, Docker, and Helm. This is not a toy project. It is a full production system with authentication, memory, retrieval, orchestration, observability, and secure service-to-service communication, all deployed on real AWS infrastructure.

What you will build and learn:

Agentic AI patterns: chaining, routing, parallelization, orchestrator-worker, and evaluator-optimizer workflows using LangGraph and Strands AgentsRetrieval-Augmented Generation (RAG) with Bedrock Knowledge Bases, OpenSearch vector search, and a dedicated Retrieval GatewayMulti-agent systems with delegation, tool use, function calling, and Model Context Protocol (MCP)LLM Gateway architecture: model routing, abstraction, streaming, and cost control across Large Language ModelsMemory and state management with PostgreSQL (Aurora), Redis (ElastiCache), and persistent agent memory

Observability and monitoring using OpenTelemetry, AWS X-Ray, and CloudWatch for full trace visibility across agents

Infrastructure as Code: provision and deploy everything with Terraform and Kubernetes (EKS) using production Helm charts

Prompt engineering fundamentals: chain-of-thought, few-shot examples, and structured evaluation techniques

What makes this course different:

You will not just copy code. Every architectural decision is explained, why each service exists, what trade-offs were made, and how the components fit together. You will underst

Prices, discounts and availability are set by Udemy. We may earn a commission when you purchase through links on this site.