The plugins allow you to extend the capabilities of Microsoft 365 Copilot, through the integration between data in the Microsoft ecosystem and information from internal business systems or third-party services such as Salesforce, Slack and GitHub. In this article, we'll dive into what Copilot plugins are, what advantages they offer, and how to create them within Microsoft Copilot Studio, the new Microsoft Power Platform development platform. Finally, we'll look at some useful use cases to understand how to use Copilot plugins to optimize your company's internal processes.
Plugins for Microsoft 365 Copilot are programs that can extend the functionality of Microsoft's new AI based on the specific needs of a company or individual user.
Generally, plugins are used to integrate data from external systems into Copilot to Microsoft 365 and which AI could not access independently. An example is the data contained in:
The advantage is that of create a unified experience, making available to Copilot all the useful data to support users in their daily activities, through the use of their words.
In addition to plugins, you can also use Microsoft Graph connectors to extend the functionality of Microsoft 365 Copilot. However, The use of plugins is preferred if the extension was used to access in real time to:
The plugins also allow you to access and use information both in reading and writing mode (Read&Write). This means that it is not only possible to view the data, but also to modify it or add new information depending on the permissions and functionality of the plugin itself.
Unlike Microsoft Graph connectors, plugins therefore allow users to interact more fully with company information through the Copilot chat and to make changes directly from their work apps, without resorting to external tools or complex procedures.
There are two approaches to developing plugins useful for extending the functionality of Microsoft 365 Copilot:
In the first case, it is necessary to rely on specialists with technical skills able to develop customized plugins starting from tools such as Teams Toolkit, Visual Studio Code and Visual Studio. You can ask intranet.ai for support to build your custom plugins through the contact page.
In the second case, Microsoft provides development tools that are more intuitive and accessible even to less technical users. Microsoft Copilot Studio is the main example, and represents the ideal solution for creating your own Microsoft 365 Copilot plugins without writing code.
Obviously, low-code tools like Copilot Studio offer limited functionality for creating Copilot extensions. If you need a personalized experience, creating custom plugins through code development is definitely the preferred way.
Copilot Studio is the new platform that brings together the Power Virtual Agents capabilities of the Microsoft Power Platform, together with the natural language processing capabilities of Azure OpenAI, to simplify programming and development activities.
To access Microsoft Copilot Studio, you must purchase one of the following licenses:
After purchasing the right license for your needs, you must meet the following requirements in order to create and distribute custom plugins:
Once these requirements are met, you can proceed with developing plugins in Copilot Studio.
The platform provides advanced features to create two different types of plugins:
The conversation plugins are therefore used to facilitate dialogue and the exchange of information between users and Microsoft 365 Copilot, while the AI plugins focus on adding artificial intelligence capabilities to analyze, interpret and use all the data of interest.
AI plugins fall into four main categories:
So let's see how to develop each of these plugins in Microsoft Copilot Studio.
In collaboration with the group company intranet.ai, Dev4Side Software offers a comprehensive adoption program to maximize investment in Microsoft 365 Copilot with the continuous support of our experts.
We are responsible for configuring the digital workplace and training users on the use of the new AI in Microsoft 365 applications. You can customize the plan according to your needs, choosing from the following services:
Conversational plugins are custom topic files that are integrated into the Microsoft 365 Copilot chat to improve the exchange of information with users.
An example would be a plugin that allows Copilot to answer questions posed by a customer to the company's Customer Service, retrieving information from the documentation stored in the customer request management system in use in the company and outside its Microsoft 365 environment.
Here are the steps to follow to create a conversational plugin for Microsoft 365 Copilot:
Once the conversational plugin has been published on Copilot Studio, a plugin is automatically created in the Dataverse registry. At this point, it is enough for the tenant admin to approve the plugin to make it immediately available in the Copilot catalog.
Request plugins fall into the AI plugin category.
Sono plugins based on the GPT model trained through Microsoft AI Builder to perform specific tasks. They are therefore able to make Copilot's answers more relevant based on the type of request made by users and on which the AI has been trained.
Copilot could be based on such a plugin for manage support requests, converting individual user prompts into requests that are addressed to members of the relevant IT department.
Here are the steps to follow to create an AI Builder-based plugin for Microsoft 365 Copilot:
The tenant admin can enable the new plugin in chats with Copilot. Just log in to the Microsoft 365 admin area and deploy the plugin found in 'Settings' > 'Integrated apps' > 'Available apps' > 'Microsoft Copilot Studio'.
Flow plugins are a type of AI plugin that focuses on performing specific actions in response to certain events. It is therefore a question of customized automations that integrate Power Automate flows with Microsoft 365 Copilot.
An example would be a plugin for automate the management of purchases.
By sending a purchase request, Copilot would be able to assign the request to a user, generate a purchase order and send update notifications through Teams, integrating data from an ERP system or a custom company application.
Here are the steps to follow to create a Power Automate-based plugin for Microsoft 365 Copilot:
Again, the tenant admin can distribute the new plugin from the Microsoft 365 admin area.
Connector plugins have a particular focus on integrating Microsoft 365 Copilot with third-party systems and services. For example, a company could develop a connector plugin for integrate Salesforce information into Microsoft Teams. The Sales department could then obtain answers and generate content for customers based on this data, directly from the Copilot chat in Teams.
Here are the steps to follow to create a plugin based on Power Platform connectors for Microsoft 365 Copilot:
As always, it will be up to the tenant admin to enable and distribute the new plugin from the Microsoft 365 administration area.
OpenAI plugins use the connection to custom OpenAI models to access information such as:
Depending on how the plugin is configured, end users can ask Microsoft 365 Copilot not only to access this information, but also to update, add, or delete content.
A company could then use an OpenAI plugin to analyze sentiment towards a communication campaign on social media, interpreting the reactions and comments published continuously for each content and generating a report useful for Marketing or Customer Service strategies.
Here are the steps to follow to create a plugin based on OpenAI templates for Microsoft 365 Copilot:
The new plugin is thus generated, together with the customized connector to be used in the company.
The Modern Apps team specializes in development and integration across the entire Microsoft 365 ecosystem. We design native applications for Microsoft and Azure platforms, and implement business processes that integrate with and maximize the investment in Microsoft 365.