Information-on-Demand with Microsoft Teams Chatbot

Microsoft’s Power Virtual Agents facilitates businesses to respond rapidly to the needs of their employees and customers. It is a low-code, GUI-driven solution that empowers your team to create rich, conversational chatbots that easily integrate with the Teams platform.

In this post, we are going to demonstrate a working chatbot that can provide simple answers to common questions.

Licensing

Here is a big surprise! If you have a Microsoft Teams license, you don’t have to pay anything extra!!!

The basic features required to implement and deploy the Microsoft Power Virtual Agent within the Microsoft Teams environment are already included in your Teams license.

So, if you are using any Microsoft 365 plan, such as the Microsoft 365 Business Basic, Business Premium, or Microsoft 365 E1, E3, or E5, the ability to use the Power Virtual Agents inside Microsoft Teams is already included in your license.

There are some limitations, but those limitations should not affect your ability to use this feature in most cases.

The Microsoft Power Platform is the de facto interface to your data inside the Microsoft 365 ecosystem.

To create and deploy the chatbot, we use these four components:

• Dataverse – this is where our data resides
• Power Apps – this is the interface we use to query the data
• Power Automate – this facilitates the flow of data between the Dataverse, Power Apps, and Power Virtual Agents
• Power Virtual Agents – the bot framework that we use to create and deploy our chatbot

The Process

The entire process of creating and deploying the chatbot is accomplished inside the Microsoft Teams environment.

Here are the most important steps:

• Create the required table and import the data into the Microsoft Dataverse (or use SharePoint List)
• Enable Power Virtual Agent and Power Apps in Teams
• Start the bot-building process using the Power Virtual Agent experience inside the Microsoft Teams
• Connect to the Dataverse and query the data using Microsoft Power Automate flow, also inside the Microsoft Teams environment
• Return the output to the chatbot

The process is easy, but it does require a technical background, and thus it is not practical to walk you through the entire process in a single post like this. If you want a complete walkthrough, please check the links in the Useful Links section.

Examples

We have named our chatbot XT1. It has built-in natural language processing. Members of the Team can interact with this bot and get answers to simple questions. For example, the bot can fetch the GST number, PAN (Permanent Account Number), address, etc. of any client in our system.

You can see the bot in action in the Animated GIF below:

Chatbots can do a lot more than what you see in these examples. For example, we can have a chatbot that can fetch “Top 5 receivables” from your accounting system with the help of Power BI.

Important Factors to Consider

• The Microsoft 365 license allows 2 GB of space per environment; this is enough for most use cases, but if you use binary objects such as images and PDFs in your flows, you will quickly run out of space
• Each tenant can have a maximum of 5 environments, plus one additional environment for every 20 licensed users; so, if you have 40 licensed users, you can have up to 7 environments and up to 14 GB of data inside Dataverse
• There is no limit on the number of bots you can create, but you can deploy the bot only inside Microsoft Teams
• If you want to deploy bots in other environments outside of Microsoft Teams (such as a website), you must upgrade to a premium license
• The bot can fetch data from various data sources in addition to the Dataverse; however, external data sources such as SQL Server require a premium license
• You can buy additional storage only if you upgrade to a premium license