custom-engine-agent

Instructions for Custom engine copilots with​ Teams AI Library and VSCode

This demo showcases how to build a custom engine copilot with Retrieval-Augmented Generation for sophisticated question answering using Teams AI library and Teams Toolkit.

Video instructions are available here.

Pre-requisites

• Node.js, supported versions: 16, 18
• Teams Toolkit for Visual Studio Code version 5.10.1
• Prepare your own Azure OpenAI, Azure AI Search and storage account resources.
• A Microsoft 365 Account with a permission to Upload custom apps to Teams.

Get started with Azure OpenAI

1. Start your demo in Azure OpenAI resource and navigate to Azure AI Foundry.
2. Create a deployment model, From the Deployments tab, select Create a new deployment. Fill out the following details and select Create:
   • Deployment name: Recommended to use the same name with the selected deployment model, such as gpt-4.
   • Select a model: Select gpt-4 or higher model.
   • Model version: Auto update to default.
   • Deployment type: Standard.
   • Content Filter: Default.
3. Create an embeddings deployment model, select Create a new deployment. Fill out the following details and select Create:
   • Select a model: text-embedding-ada-002.
   • Model version: Default.
   • Deployment type: Standard.
   • Deployment name: Choose a memorable name, such as text-embeddings
   • Content Filter: Default.
4. Once your models are successfully created, you can navigate to Chat. In the Setup section, select Add your data tab and then Add a data source.
5. Select Upload files (preview), then fill the details as the following and select Next:
   • Subscription: Select the subscription you created your Azure resources.
   • Select Azure Blob storage resource: Select your storage resource. (You'll see a message Azure OpenAI needs your permission to access this resource, select Turn on CORS.)
   • Select Azure AI Search resource: Select your Azure AI Search resournce.
   • Enter the index name: Index name, such as resumes; make note of this.
   • Check the box for Add vector search to this search resource and select your embedding model. Then select Next.
   • Select Browse for a file and select the pdf documents from the data folder. Then, select Upload files and Next.
   • Select Search type as vector and chunk size as 1024(Default), then Next.
   • Select API Key as Azure resource authentication type, then Next.
6. Once your data ingestion is completed, use Chat playground to ask questions about your data such as:
   • "Suggest me .NET developers who can speak Spanish"
   • "Suggest me python developers in London Area"

Bring your data to Teams

1. Clone this github repo, navigate to custom-engine-agent folder and open it in Visual Studio Code.
2. In the project folder, select env folder and rename:
   • .env.local.sample file to .env.local
   • .env.local.user.sample file to env.local.user
3. Open env.local.user file and fill the values as shown below and save the file:

SECRET_BOT_PASSWORD= <LEAVE BLANK>
SECRET_AZURE_OPENAI_API_KEY='<YOUR-AZURE-OPENAI-RESOURCE-KEY>'
AZURE_OPENAI_ENDPOINT='<YOUR-AZURE-OPENAI-RESOURCE-ENDPOINT>'
AZURE_OPENAI_DEPLOYMENT_NAME='<YOUR-AZURE-OPENAI-RESOURCE-MODEL-NAME>'
AZURE_OPENAI_EMBEDDING_DEPLOYMENT_NAME='<YOUR-AZURE-OPENAI-RESOURCE-EMBEDDINGS-MODEL-NAME>'
SECRET_AZURE_SEARCH_KEY='<YOUR-AZURE-AI-SEARCH-KEY>'
AZURE_SEARCH_ENDPOINT='<YOUR-AZURE-AI-SEARCH-ENDPOINT>'
INDEX_NAME='<INDEX-NAME>'
TEAMS_APP_UPDATE_TIME= <LEAVE BLANK>
SECRET_AAD_APP_CLIENT_SECRET= <LEAVE BLANK>
HR_EMAIL= <YOUR-DEMO-TENANT-ACCOUNT-EMAIL>

Select Teams Toolkit on Visual Studio Code navigation bar, login with Microsoft 365 Account.

Make sure that sideloading is enabled for your account. If Sideloading is not enabled yet:

1️⃣ Navigate to https://admin.microsoft.com/, which is the Microsoft 365 Admin Center.

2️⃣ In the left panel of the admin center, select Show all to open up the entire navigation. When the panel opens, select Teams to >open the Microsoft Teams admin center.

3️⃣ In the left of the Microsoft Teams admin center, open the Teams apps accordion. Select Setup Policies, you will see a list of >App setup policies. Then, select the Global (Org-wide default) policy.

4️⃣ Ensure the first switch, Upload custom apps is turned On.

Let's test the demo on Teams. Start debugging your app by selecting Run and Debug tab on Visual Studio Code and Debug in Teams (Edge) or Debug in Teams (Chrome). Microsoft Teams will pop up on your browser. Once your app details show up on Teams, select Add and start chatting with your app.

Tip: Testing this exercise locally

Make sure to test and debug this exercise on Teams locally, as some of the Teams AI library capabilities you've implemented in >your app so far won't smoothly work in the Teams App Test Tool.

Ensure your questions are related to your dataset. Go through pdf documents in the resumes folder to understand more about your data. Challenge your custom engine agent by combining requirements and asking complex questions! Some suggestions would be:

Can you suggest a candidate who is suitable for spanish speaking role that requires at least 2 years of .NET experience? Who are the other good candidates? Who would be suitable for a position that requires 5+ python development experience? Can you suggest any candidates for a senior developer position with 7+ year experience that requires Japanese speaking?