developting a production level RAG workflow prompt flow Hello Anonymous 👋🏽.

This is a proctored 75-min lab to on “Developing a production-level RAG Workflow” with Azure AI Studio & Prompt Flow.

Take a minute to familiarize yourself with the instructor and proctors for this session. The lab is organized into the sections below, with numbered steps in each. Use the section and step number as context for proctors when asking questions. Pre-requisites

You must have the following to participate in this lab: GitHub Account - with GitHub Codespaces. Free quota is sufficient. Your own laptop - fully-charged. This is a 75-minute lab. Modern browser - on laptop. To launch the Lab-on-Demand session. Table Of Contents

Lab Overview

  1. Get Started
  2. Launch GitHub Codespaces
  3. Verify Azure Is Provisioned
  4. VSCode Azure Login
  5. VSCode Azure Config
  6. VSCode Config Env
  7. VSCode Populate Search
  8. VSCode Populate Database
  9. VSCode Config Connections
  10. Azure Config Connections
  11. PromptFlow Explore Codebase
  12. PromptFlow Open Visual Editor
  13. PromptFlow Run Flow
  14. PromptFlow Evaluate Flow
  15. Push PromptFlow To Azure
  16. PromptFlow Deploy Flow Lab Recap Appendix

Lab Overview

This lab gives you hands-on experience with End-to-End LLM Application Development (LLMOps) by teaching you to build, run, evaluate, and deploy a RAG-based application (“Contoso Chat”) using Azure AI Studio and Prompt Flow.

Learning Objectives

By the end of this lab, you should be able to:

Explain LLMOps concepts & benefits. Explain Prompt Flow concept & benefits. Explain Azure AI Studio features & usage. Use Prompt Flow on Visual Studio Code Design RAG-based LLM Applications Build, run, evaluate & deploy RAG-based LLM apps on Azure. Pre-Requisites

The lab environment is pre-configured with an Azure subscription and pre-provisioned with the required Azure resources to jumpstart your journey. We assume some familiarity with the following concepts:

Machine Learning & Generative AI concepts Python & Jupyter Notebook programming Azure, GitHub & Visual Studio Code tooling Development Environment

You’ll use the following resources in this lab:

Contoso Chat - as the target application. Github Codespaces - as the dev container Visual Studio Code - as the default editor Azure AI Studio (Preview) - for AI projects Azure ML Studio - for minor configuration Azure Portal - for managing Azure resources Prompt Flow - for streamlining end-to-end LLM app dev