Our Blog

where we write about the things we love



Get fancy with a Bot – Microsoft Dynamics 365 Virtual Agent explained

Would you like to empower your team to deliver better service and boost your customer satisfaction? If the answer is yes, we have some good news for you. You can now easily create powerful bots using a guided, no-code graphical interface.

The Microsoft Dynamics 365 Virtual Agent for Customer Service is now in public preview and available in the US, with global release pending. At Intergen, we live and breathe Dynamics 365 and have been experimenting with Virtual Agent to see how it could help you boost your customer engagement.

Here’s our guided tour…

What is Microsoft Dynamics 365 Virtual Agent for Customer Service?

Virtual Agent for Customer Service is a software-as-a-service (SaaS) offering. It allows you to easily register, create your chat bot, and embed it into your website with just a few clicks. There is no infrastructure to maintain, no complex systems to deploy.

This bot is built on the Microsoft Bot Framework and has natural language capability. It can be built and deployed by subject matter experts to automate common support enquiries. What are your business opening hours? Who should the customer speak to for customer support? When is their next bill due? What steps should the customer take to troubleshoot product-related issues?

The answers to these questions can easily be automated, allowing customer service staff to focus on other complex issues and enquiries, lowering the workload of mundane and repetitive cases.

What does it offer?

With Microsoft Dynamics 365 Virtual Agent for Customer Service you get the following:

  • A no-code solution that empowers your customer service teams by allowing them to easily build bots themselves without needing intermediaries, or coding or AI expertise.
  • Automated responses to low-level enquiries that helps you increase customer satisfaction and reduce costs, by providing quick answers to common questions and free your human agents up to deal with more complex issues.
  • 24/7 operation that allows customers to get self-help whenever it suits them, without having to wait for support desk opening hours.
  • Integration with data sources using hundreds of easy-to-add custom connectors via Microsoft Power Automate (previously Microsoft Flow), to help you build effective customer service processes and draw on your data sources to deliver meaningful answers.

Why Virtual Agent is different to other bots?

  • Easy to build and deploy: It allows your non-technical Subject Matter Experts to monitor the bot using built-in analytics, and to update and enhance the bot using an online web interface.
  • No expertise needed: It removes the complexity of exposing teams to the complexities of conversational AI and the need to write complex code. It minimises the IT effort required to deploy and maintain a custom conversational solution.
  • Enable rich, natural conversations: Microsoft’s powerful conversational AI capabilities enable end users to have rich multi-turn conversations that quickly guide them to the right answers. There’s no need to retrain AI models.
  • Monitor and improve bot performance: Virtual Agent for Customer Service lets you keep an eye on how your bots are performing using powerful metrics and AI-driven dashboards. Easily see which topics are doing well, where the bot can improve, and quickly adjust improve performance.

Getting started!

Let us assume you are working as part of the customer service team of an aged care facility. The team is busy answering the same questions repeatedly. Why don’t you automate how enquiries are received from your public website?

You create a guide experience using the chat bot, prompting website visitors to submit the relevant information.The information submitted then creates an enquiry record in Dynamics 365 Customer Engagement (CE) which can be assigned to the right team member to follow-up.

The following steps outline the high-level process involved in setting up a bot for the above scenario:

Get fancy with a Bot – Microsoft Dynamics 365 Virtual Agent explained

Create your first bot

Once you have provisioned an environment you can easily create your first bot by following these simple instructions.

Create custom topics for your bot

After creating your bot, you can create your own custom topic - or, in other words, a conversation tree specifying how your bot responds to a user’s question.

A topic defines a conversation path with the bot for a specific customer intent. In the topic we can specify a trigger phrase that prompts the bot to start the conversation.

Dynamics 365 Virtual Agent for Customer Service includes built-in topics to be used as building blocks for developing the bot's conversation path. The built-in topics can give you a head start on developing your bot.

Follow the instructions here for steps to create custom topics. For the above example I have created two custom topics:

a) Facility information: This topic helps a customer with a guided experience while enquiring for information about different facilities. It provides the customer with information for the facilities depending on the region they are interested in. It has multiple branches offering the various regional facility information.

b) Submit an enquiry: Using the guided experience, the customer is further prompted to submit an enquiry. Once the customer confirms they are enquiring about the right facility, they can submit their enquiry. This topic contains the flow action that will submit an enquiry record to a Common Data Service (CDS) instance.

Connect the topics

Different topics can be connected using the ‘Go To’ action to build a conversation path. This way, topics can be reused reducing the time to create individual conversation paths. Virtual Agent provides some built-in topics which can be used as building blocks for developing the bot's conversation path.

Create Flow to submit the enquiry into the CDS

You can enable your bot to perform an action by invoking a flow.

Important: Use a flow that shares the same CDS environment as the bot.

Verify that the PowerApps environment database was created correctly. Verify that the Solutions list includes Common Data Service Default Solution. The flow should be created in this solution for it to appear in the ‘Actions’ in the authoring canvas of the topic.

Once the flow is created it will appear in the authoring canvas pane of topic under ‘Actions’.

Test your bot

As you design your bot in Dynamics 365 Virtual Agent for Customer Service, it can be tested in real time using the ‘Test bot’ pane. This shows how the bot guides a customer through the conversation path.

This helps you to find and fix unexpected behaviour, and you can enable tracing to take you through the conversation path step by step. You can then go to the corresponding node in the conversation editor and rectify it.

Deploy your bot

Once you have finished designing your virtual agent, you can deploy it to a demo environment on the web or to your live website. Find the deployment instructions here.

Use analytics to improve your bot

Once your bot has completed interactions with users, the resulting statistics are available via the Analytics tab in the left pane. Here, you can find key performance indicators (KPIs) showing the volume of sessions your bot has handled, how effectively your bot was able to engage end-users and resolve issues, escalation rates to human agents, and abandonment rates during conversations.

You will also find customer satisfaction information at the KPI level as well as in the Customer Satisfaction tab.

Let’s Talk

Convinced of the value of deploying Virtual Assistant? Talk to Intergen to learn more about how you can empower your teams to be more effective by sharing the workload with a virtual team member.


This blog is part of the #cxreimagine series. For more experts' insights, clients' experiences and to download the whitepaper, click the banner.

For more experts' insights, clients' experience and to download the whitepaper, click the banner #cxreimagine

Posted by: Monica Chopra, Senior Consultant | 02 December 2019

Tags: Bot Framework, #CXreimagine

Top Rated Posts

Blog archive

Stay up to date with all insights from the Intergen blog