Reference Architectures

Three architectures that cover most LLM systems — the managed-API app, the RAG stack, the self-hosted cluster — drawn end to end, per cloud.

Content last verified 2026-09.

Lessons

1. The Managed-API App
2. The RAG Stack
3. The Self-Hosted Cluster
4. Choosing and Evolving

Sources

• Amazon Bedrock User Guide
• Azure AI Foundry documentation (Microsoft Learn)
• Vertex AI documentation