Deploy Web Application to Interact with OpenAI and Cognitive Search Demo

Attribution

• The code in this repo is adapted from the Azure Samples ChatGPT + Enterprise data with Azure OpenAI and Cognitive Search Demo

Features

• Deploy a python based web application interface to an Azure App Service Environment that interacts with Azure OpenAI, Cognitive Search, and Azure Storage Blob.

Getting Started

IMPORTANT: In order to deploy and run this example, you'll need an existing Azure subscription with deployed Azure OpenAI service, Cognitive Search, Azure Storage Account and documents loaded in per the Azure Samples ChatGPT + Enterprise data with Azure OpenAI and Cognitive Search Demo.

AZURE RESOURCE COSTS by default this sample will create Azure App Service that have a monthly cost.

Prerequisites

For Deployment to Existing App Service Environment

• Azure CLI
• Bicep

To Rebuild the Web App Source Code

• Azure Developer CLI
• Python 3+
  • Important: Python and the pip package manager must be in the path in Windows for the setup scripts to work.
  • Important: Ensure you can run python --version from console. On Ubuntu, you might need to run sudo apt install python-is-python3 to link python to python3.
• Node.js
• Git

NOTE: Your Azure Account must have Microsoft.Authorization/roleAssignments/write permissions, such as User Access Administrator or Owner.

Installation

1. Configure Environment Variables or Update main.parameters.json file with apprpriate values.
2. Deploy the App Service Plan and App Service from the Bicep file

az deployment sub create --location <location> --template-file infra\main.bicep --parameters infra\main.parameters.json

3. Upload the web application code

az webapp deploy --name <app service name> --resource-group <resource group name> --src-path app-deploy.zip

Rebuild Web App and Redeploy

1. Run the azd package command

azd package

The output of successfully running the above command will show the path where the new .zip package will be located. 2. Upload the generaged zip package to your app service

az webapp deploy --name <app service name> --resource-group <resource group name> --src-path <path/to/recompiled-app.zip>

Quickstart

• In Azure: The URL is printed in the outputs when the bicep template completes (as "BACKEND_URI"), or you can find it in the Azure portal by browsing to the Application Service that was deployed.

Once in the web app:

• Try different topics in chat or Q&A context. For chat, try follow up questions, clarifications, ask to simplify or elaborate on answer, etc.
• Explore citations and sources
• Click on "settings" to try different options, tweak prompts, etc.

Resources

• Revolutionize your Enterprise Data with ChatGPT: Next-gen Apps w/ Azure OpenAI and Cognitive Search
• Azure Cognitive Search
• Azure OpenAI Service