AWS AI Reference Architectures

Six production-shaped reference architectures for AI workloads on AWS — diagrams, decisions, Terraform skeletons, cost analysis and Well-Architected reviews.

This is the doc site for github.com/fernandofatech/aws-ai-reference-architectures. Every architecture answers the same questions in the same order, so you can compare designs at a glance.

The architectures

#	Architecture	Pattern	Best for
01	RAG with Bedrock + OpenSearch	Retrieval-augmented generation	Internal Q&A over docs
02	Multi-agent orchestration	Bedrock Agents + Step Functions	Long-running workflows
03	Streaming AI inference	API Gateway + Lambda streaming	Chat UIs with token streaming
04	Event-driven AI processing	EventBridge + SQS + Lambda	Async classification & enrichment
05	Fine-tuning pipeline	SageMaker + S3 + MLflow	Custom models on top of FMs
06	Secure agentic system	Bedrock Agents + Guardrails + VPC	Multi-tenant production agent

Each page follows the same eight-section template. Read just the trade-offs if you are time-constrained — it's the section that tells you whether the pattern applies to your problem.

Author

Fernando Francisco Azevedo — Solution Architect, AWS & AI focus. fernando@moretes.com · LinkedIn · fernando.moretes.com