An Introduction to Power Virtual Agents for Microsoft Teams

Posted on by

Over the past few years, it has become increasingly easier to create powerful chatbots with little to no code. Now almost anyone can quickly create a deploy a chatbot with ease. Power Virtual Agents is a premium add-on to the Power Platform that allows you to create and deploy such bots.

However, as part of Dataverse (previously called Project Oakdale) many Microsoft 365 subscribers have access to Power Virtual Agents for Microsoft Teams. These chatbots are covered under their Microsoft 365 licensing and require no premium licensing to develop, deploy, or use.

There are limitations with the Power Virtual Agents for Microsoft Teams. Unless you upgrade to a full Power Virtual Agents subscription you will not be able to:

  • Deploy chatbots to a specific Channel or specific Team.
  • Utilize premium connectors in Power Automate
  • Access and Update the source code for the bot using the Bot Framework

You can find out more about licensing for Power Virtual Agents at https://docs.microsoft.com/en-us/power-virtual-agents/requirements-licensing-subscriptions

Don’t let this dissuade you though! Even with these limitations, users can build exceptionally powerful chatbots with little to no effort with an intuitive interface. Using Power Virtual Agents for Microsoft Teams users can:

  • Create and deploy bots to the left rail in Teams that other users in your organization can take advantage of
  • Use Power Automate flows to connect to other systems, send emails, and perform other business logic
  • Create bots that prompt users with follow up questions, use conditional logic, and help guide the user to perform their tasks
  • Review and analyze the performance of the bot, such as what questions are being asked the most, and how effective the bot is at understanding and answering questions

It’s really amazing what you can accomplish. Check out the following video to show you how quickly and easily you can get started building chatbots with Power Virtual Agents for Microsoft Teams.

To recap some of features I show you in the video:

  • Users create Power Virtual Agents by using the “Power Virtual Agents” application located in Teams
  • A “Topic” is a specific job or conversation thread that bot executes. Topics have “trigger” phrases to determine if a user is wanting to perform a certain task. For example, “Hello” and “Hi” would trigger the “Greeting” Topic.
  • There are a handful of default topics already created that users can customize as needed (ie: Greeting)
  • You can create actions (Power Automate Flows) where data can be passed into the flow and returned back to the Power Virtual Agent.
  • Using the Authoring Canvas, you can create conditional branches and prompt users for more information (check out the pre-baked “Lesson 3” Topic when you create a bot for good example).
  • You can pass the Azure ID of the user interacting with the bot to Power Automate in order to retrieve specific information about the user (in the video I retrieve the current user’s Manager)
  • If you run into issues while testing, be sure to “Reset” your bot.
  • When you update your bot, it may be necessary to start a conversation with “Start over” to get your updated changes.

All in all, Power Virtual Agents for Microsoft Teams is one of my new favorite things and I look forward to digging into it more. Check it out! You might surprise yourself with what you can build!

To find out more about Power Virtual Agents for Microsoft Teams you can visit: https://docs.microsoft.com/en-us/power-virtual-agents/teams/fundamentals-what-is-power-virtual-agents-teams

I’d love to hear your thoughts on Power Virtual Agents for Microsoft Teams and how you are taking advantage of them in your organization!

https://techcommunity.microsoft.com/t5/microsoft-teams-community-blog/an-introduction-to-power-virtual-agents-for-microsoft-teams/ba-p/2049485 https://techcommunity.microsoft.com/t5/microsoft-teams-community-blog/an-introduction-to-power-virtual-agents-for-microsoft-teams/ba-p/2049485 2021-01-11 16:54:56Z