Enterprise RAG architecture on OCI integrating Object Storage, OpenSearch vector database, and Generative AI Service for semantic search and LLM inference. Data flows from ingestion through ETL pipelines into embeddings, retrieval via OpenSearch, and orchestrated inference through OCI Compute and Functions. Multi-subnet VCN design isolates data ingestion, retrieval, and inference workloads with API Gateway, Load Balancer, and WAF protecting user-facing endpoints. Security and governance layers including Vault key management, IAM, Cloud Guard, and comprehensive logging ensure compliance and auditability. Fork this diagram to customize subnets, add additional compute nodes, or integrate with your existing OCI tenancy.