1.4 RAG Chat Tutorial¶

This markmap provides the big picture for navigating this RAG Chat tutorial. Learn to build a minimal RAG-based copilot experience using Azure AI Foundry Portal (for setup) and Azure AI Foundry SDK (for ideation and evaluation).

RAG Chat Roadmap: Refresh this page to see markmap if needed

# RAG Copilot ## 1. Overview ### 1.0 Pre-Requisites - Azure Subscription (Roles) - GitHub Account (Codespaces) - AI Models (Chat, Embedding) - Application Data (RAG) ### 1.1 Concepts - Generative AI Ops (GenAIOps) - Custom Copilot (Chat AI) - Prompt Template (Asset Format) - Retrieval Augmented Generation (RAG) - AI-Assisted Evaluation (LLM-As-Judge) - [Azure OpenAI Deployment types](https://learn.microsoft.com/en-us/azure/ai-services/openai/how-to/deployment-types#global-standard) - [Data Residency in Azure](https://azure.microsoft.com/en-us/explore/global-infrastructure/data-residency/) ### 1.2 Technologies - Azure AI Foundry Portal - Azure AI Project Resource - Azure AI Hub Resource - Azure AI Search Service - Azure OpenAI Service ### 1.3 Dev Tools - GitHub Codespaces (Dev Env) - Visual Studio Code (Dev IDE) - Azure CLI (Configure) --- ## 2. [Setup](https://learn.microsoft.com/en-us/azure/ai-studio/tutorials/copilot-sdk-create-resources) ### 2.1 Create AI Project ### 2.2 Deploy AI Models ### 2.3 Add Azure AI Search ### 2.4 Setup Local Environment ### 2.5 Configure Env Variables ### 2.6 Validate Env Setup ### 2.7 Connect The Dots - What did we do? - Why did we do it? - How can we improve? --- ## 3. [Ideate](https://learn.microsoft.com/en-us/azure/ai-studio/tutorials/copilot-sdk-build-rag) ### 3.1 Add Application Data ### 3.2 Create Search Index ### 3.3 Retrieve Related Products ### 3.4 Extract Customer Intent ### 3.5 Retrieve Related Knowledge ### 3.6 Design Grounded Prompt ### 3.7 Validate the Prototype ### 3.8 Connect The Dots - What did we do? - Why did we do it? - How can we improve? ## 4. [Evaluate](https://learn.microsoft.com/en-us/azure/ai-studio/tutorials/copilot-sdk-evaluate) ### 4.1 Create Evaluation Dataset ### 4.2 Create Evaluation Script ### 4.3 Configure Evaluation Model ### 4.4 Run Evaluation Script ### 4.5 View Results Locally ### 4.6 View Results in Portal ### 4.7 Validate Evaluation ### 4.8 Connect The Dots ## 5. Evolve ### 5.1 Recap: Build a Copilot ### 5.2 Refactor: Make it Better ### 5.3 Resources: Learn More