Contoso Support Chat Prompt Flow Demo

This repository hosts the code used in the MS Ignite 2023 | Live Breakout Session | BRK203

End-to-End AI App Development: Prompt Engineering to LLMOps

Introduction

Prompt engineering and LLMOps are pivotal in maximizing the capabilities of Language Models (LLMs) for specific business needs. This code base is from the End-to-End AI App Development: Prompt Engineering to LLMOps Microsoft Ignite session. We'll walk you through the entire process—from prototyping and experimenting to evaluating and deploying your AI-powered apps. Learn how to streamline your AI workflows and harness the full potential of Generative AI.

Prerequisites

• Azure Subscription
• VS Code
• Prompt Flow Extension
• anaconda or venv to manage python environments.
• Currently this app needs to be run using WSL. To install WSL follow the instructions here.

Getting Started

Clone the repo

git clone https://github.com/microsoft/contoso-support-pf-demo.git

Open the repo in VS Code

cd contoso-support-pf-demo
code .

Install the Prompt Flow Extension

• Open the VS Code Extensions tab
• Search for "Prompt Flow"
• Install the extension

Create a new python environment

Using anaconda

conda create -n contoso-support-pf-demo python=3.8
conda activate contoso-support-pf-demo
pip install -r requirements.txt

Using venv

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Create data resources to be used in the prompt flow

• Azure AI Search - Create an Azure Cognitive Search service The Azure AI Search connection is used to do vector lookups and return the information from the Azure AI Search index. The Azure AI Search index is created from the data in the data/product_info folder with the script\create_index.sh.

• Azure Open AI Connection - Create an Azure Open AI Connection The Azure Open AI connection is used to create the endpoints for GPT-3 and GPT-4 as well as the embeddings model.

• Azure Cosmos DB - Create an Azure Cosmos DB The Azure Cosmos DB connection is used to store the customer data needed to lookup the customer information and return the information to be populated in the prompt.

Setup the Connections

To run the prompt flow, the connections need to be set up. These can be setup as local connections or with the json confirguration connected to your workspace.

• To setup local connections follow the instructions here
• To setup json connection follow the insturctions here

Building a Prompt flow

Now that the environment, extensions, and connections have been installed we can open up the prompt flow and take a look at what it does.

• Click on the flow.dag.yaml file in the explorer. If everything was installed and the python environment was activated you should see the following and select visual editor to view the propmt flow:

• When the flow opens you should see the following:

• To learn how to build this flow from scratch, check out the session and Microsoft Ignite.

Running Chainlit App Evaluation Tool

Once everything has been setup you can run the evaluation tool to evaluate the prompt flow. To run the evaluation tool, run the following command in the terminal:

cd contoso-support-pf-demo/support-retail-copilot
chainlit run exp/app.py -w

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.opensource.microsoft.com.

When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.

Trademarks

This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party's policies.