18 Aug
Creating Intelligent Chatbots Using Power Virtual Agents & Dataverse
Over the past two years, the shift in how businesses operate worldwide has been palpable. Dynamics have transformed, ushering in an era of rapidly changing customer and employee expectations.
One of the most significant trends witnessed over the past two years has been the unprecedented and unparalleled rise in the use of conversational AI — chatbots and virtual agents — toagents—to provide quick, meaningful, and more personalized service to customers while enhancing employee efficiency levels by minimizing repetitive and redundant tasks.
A well-developed chatbot has the potential to shape a company’s brand image, playing an integral role in attracting and retaining the right market for your product and keeping employees active, engaged, and motivated throughout their journey with your organization.
Agile, resilient, and future-proof — these are three words that best describe today’s business environment. It is not enough to implement chatbots; deploying them at the right time to hasten business operations and build a competitive edge in the market is paramount.
That’s where Microsoft’s Power Virtual Agents (PVA), a low-code SaaS platform that helps build and deploy bots quickly with minimum coding experience required, comes into the scene.
This blog will explore creating intelligent chatbots using Power Virtual Agents and the Dataverse.
An integral part of Microsoft’s suite of services, Power Virtual Agents helps organizations develop and implement virtual agents and chatbots leveraging the low-code approach. This means both pro and citizen developers alike can build chatbots in a short period with little or no coding knowledge needed.
Moreover, being hosted on the Microsoft Azure Cloud environment offers the added advantage of pre-existing infrastructure. Hence, organizations needn’t invest more resources in maintaining the chatbots and can enjoy seamless integration with other services.
In short, Power Virtual Agents capitalizes on its graphical interface and minimalistic coding requirements to make bot building a breeze for all.
PVA takes entire responsibility for the functional aspect of the chatbot, so all the user has to do is provide the logic and set the virtual agent up.
Below, we describe in brief the process of creating your first chatbot using Power Virtual Agents and the Dataverse environment —
- Install Power Virtual Agents in Teams by selecting and adding it from the left navigation bar.
- After installing, launch the app and choose the “Start Now” button.
- Choose which team would own and manage your chatbot and then “Continue.”
- Choose a name and language for your virtual assistant.
- Select “Create” to complete the process of creation.
- Define the topic nodes. Topics define chatbot conversations and how they take place. Types of topics include “trigger” phrases (phrases and keywords that indicate the user has asked a question) and “conversation nodes” (define how the bot responds to the user’s request).
How to create a topic?
- Select + New topic. Name and Save the Topic.
- Enter trigger phrases for the topic, “for example, “Who can I contact for a loan query?”
- In the Message node, enter information for the trigger nodes. For example, “The contact details for a loan query are Steve Robbins, sr@xyz.com.” You can add multiple conversation nodes.
- Select Redirect to another topic > End of conversation to define the end of the conversation.
Click here to know more.
7. Add inputs, variables, and conditions.
Instead of defining all trigger phrase information in the message node, you can add inputs, variables, and conditions to vary the chatbot’s response based on the user’s input.
Inputs define the user’s response when the chatbot asks a question. While variables store the inputs to be used later, conditions set forth the branching logic.
- To add a node between the trigger phrases and end conversation node, select the Add node plus sign and then select the “ask a question” option.
- Select “Multiple choice options” and define the options for the user to select from.
- Select the variable name that will be triggered when the user configures the branching logic.
- To define the next appropriate response for each of the multiple choice options, create a new condition node and add a new message node for each probable response that the user may give.
8. Pull in data from Dataverse for Teams
Dataverse for Microsoft Teams is a low-code service that offers relational data storage, editable data tables, rich data types, robust governance, and one-click deployment to Teams.
- You can create a Dataverse for teams table in the same team as the chatbot using Microsoft PowerApps.
- In PowerApps for Teams, select Build -> See all -> New -> Table.
- Create a new table, add new rows, and columns.
- Add a “call an action.”
- Use Power Automate flows to pull in data from the Dataverse for Teams table, Lists, or any other data source.
To know more, click here
9. Publish the Chatbot
- In the Power Virtual Agents app, open the chatbot for editing.
- Choose the Publish bot menu item or button.
- Choose the Publish option. Select Publish to confirm in the Publish latest content confirmation window.
10. Publish the Chatbot
Once the chatbot is complete, it must be published so that users can interact with it. You can send a link to team members or add the chatbot to the Built by your colleagues section in Teams or if the chatbot is for the organization, you’ll send it to your Teams admin for approval.
If the chatbot is only for your team members, you must instruct them on how to use and share the bot by choosing one of the following:
- Copy link – Provide users with a link to the chatbot.
- Add to the team – Add the bot to a specific team. With this option enabled, anyone in your team will be able to find the bot in the Teams app store in the Built by your colleagues section.
- Show in Teams app store – Make your bot visible in the Teams app store, with the option of showing only to teammates and shared users or to everyone in your organization.
Click here to learn about how to publish the chatbot for the entire organization.
Subscription And Licensing:
Microsoft 365 offers two editions of PVA:
- PVA for Teams: A free app that only permits the creation of internal bots.
- Power Virtual Agents: Requires additional subscription & supports the development of both web-based external client bots and internal Microsoft Teams bots.
How can Acuvate help?
As a Microsoft Gold Partner and provider of next-generation AI and consulting services, Acuvate leverages Microsoft’s robust suite of services, including PowerApps, Power Virtual Agents, and Power Automate, to create custom apps, AI bots, and workflows that enhance employee efficiency and support the delivery of exceptional customer experiences.
We have assisted clients from different industries and geographies in creating intelligent chatbots using PVA and Microsoft Dataverse.
To learn more about our chatbots, please schedule a personalized consultation with one of our experts.
Satheesh Kothakapu is Technical Architect at Acuvate and brings in 10+ year of strong expertise across Microsoft stack. He has consulted with clients globally to provide solutions on technologies such as Cognitive Services, Azure, DevOps, Virtual Agents. Currently he manages key customer engagement, involves in architecting the solutions and leading the team of Azure services.
Satheesh Kothakapu