What is Copilot Studio?
Microsoft Copilot Studio is an innovative graphical interface designed to create and manage copilots, introduced with latest updates during MS Build’24, over existing Power Virtual Agents. These copilots are AI-powered conversational interfaces leveraging Large Language Models (LLMs) and various knowledge bases. With Copilot Studio, you can deploy your copilot across multiple platforms such as web browsers, custom websites, Microsoft Teams, Microsoft 365, Facebook, Telegram, and more. You can enhance its capabilities by integrating Power Apps to monitor performance metrics. Moreover, it supports integration with diverse knowledge sources including public databases, files, SharePoint, Dataverse, and Microsoft Fabric, enriching the copilot's ability to generate informative responses and engage effectively. It also provides you with some default templates, which you can use to kick start your journey at https://learn.microsoft.com/copilot/.
Updates in Copilot Studio from Power Virtual Agent
Let's Create Our Own Copilot!
To start building your copilot, you can use either the web app or integrate directly within Microsoft Teams using Copilot Studio. We will work together to build “MLSA Guide” copilot. Here’s a quick guide to get you started:
Navigate to the Home Page
The Copilot Studio home page provides an overview of your recent copilots. From here, you can create new copilots, access recent projects, and find learning resources like documentation and tutorials
Create Your Copilot
1. Sign Up and Select Environment
2. Configure Your Copilot
Customize the Copilot
Now that you've built a copilot capable of answering questions about the topic, remember it hasn't been published yet. Before proceeding with publishing, let's delve into further modifications we can make.
Let’s Explore Different sections :-
3. Topics: When setting up a new copilot, you can begin by crafting topics. A topic serves as a structured dialogue tree that guides how your copilot interacts with users. It's designed to provide a human touch and address specific inquiries systematically. For instance, in scenarios like sales where you need to initiate an order following a defined process, these steps can be outlined within topics. Some topics are predefined in your copilot like greeting, thank you, start over, conversation boosting etc.
Let's delve into creating a topic named "Requirement to become MLSA." Which helps to guide students to know the requirements and path activities, they must select to join community. We will follow these steps to observe the outcome:
a. Click on Topic >"Add a topic" > "From Blank." You can also utilize Copilot to outline your topic by providing a description. Here, we'll start fresh with a blank template.
b. Trigger : Now we will build our topic. Start by naming your topic and configuring a Trigger. A trigger acts like a welcome phrase for your Copilot conversation. It determines when a specific topic within your Copilot should be activated based on the user asks. Click on ‘edit’ in the trigger to open ‘on recognised intent’ pane, which provides you with different properties attached with it. You can add phrases, which will appear on the right in trigger box. Once you have added all relevant trigger phrases hit ‘save’. The more phrases you include, more easily your copilot can trigger the topic.
c. Message : Now since we have initiated a Trigger, with some defined phrases which will get activated when a user seeks answer, in our scenario to the question of the type “ How to become MLSA” . Now, it’s time to configure the answer, by adding a message box attached to trigger. You can also reply using card, images, video and more.
i. Click on ‘+’ > ‘Send a message’.
ii. Now a message box appears. Add the text to it, which will answer the user’s query.
d. Question : Now, since you have answered initial questions, there are changes that user want to be more specific. In this case, you can use and add more nodes to your current topic. Like, for our copilot, I'd like to confirm whether you need more details about the process mentioned in the message. For instance, regarding the requirements for a community builder, influencer, or startup advocate. To make this clearer, I've created a Multiple-Choice Question with specific properties.
e. Variables : Variables are used to store user responses and can have various properties attached to them. You can specify the scope, value, reference, and type of a variable such as Boolean, string, number, or choice, based on your needs. Click on 'Variables' to explore them. You can select any variable you wish to modify its properties. Some variables are predefined by the system, while others are custom or environment specific.
4. Test : Now, save the topic and test your copilot. If it responds according to your specifications, then congratulations! You have successfully created your topic. If not, debug it again to identify and correct any errors.
7. Channels – In this section, you will see different options to deploy your copilot, such as Teams, Facebook, a custom website, Skype, and more.
I have deployed my copilot on teams, you can access through link – Become MLSA . Here is a snapshot of it.
Resources
Thank you for taking the time to read our blog! We hope you found it informative and helpful. If you have any questions or feedback, please feel free to leave a comment below. Don't forget to subscribe for more updates and insights. Happy deploying!
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.