Copilot Agents
In an era where efficiency and productivity are paramount, the integration of artificial intelligence (AI) into our daily workflows has become increasingly essential. Microsoft 365 Copilot has already revolutionized the way we interact with our productivity tools, offering intelligent assistance that streamlines tasks and automates routine processes. But the innovation doesn’t stop there. Enter Copilot Agents – the next evolution in AI-powered user interfaces, designed to make Microsoft 365 Copilot even more indispensable.
What is Microsoft 365 Copilot?
Let’s get the elephant out of the room to start. Microsoft has always struggled with names, dating back to when the Office team fought with the Windows folks who fought with everyone else. It has been a sea change since Satya took over – as many divisions were merged and the naming cleaned up – to mostly revolve around a singular brand: Microsoft 365.
In that spirit, Microsoft last month renamed all of their Copilots and radically changed how they were going to market. Instead of having twenty different Copilot products – each with their own branding and value proposition – Microsoft has now settled on a new positioning that everyone should have Microsoft 365 Copilot and that it will be the “UI for your AI” – the tool that everyone will be able to use – leveraging a universe of “Copilot Agents” that hook into various systems.
That means former products like “Copilot for Sales” and “Copilot for Service” have now become “Microsoft 365 Copilot for Sales” and “Microsoft 365 Copilot for Service”. A mouthful, right? The good news is that you can now simply describe each of these as a Copilot Agent, that hooks into your sales system, or your service system. Microsoft today has three first party Copilot Agents: the two above and “Microsoft 365 Copilot for Finance”. We anticipate they will continue to build additional first party agents over the next few months.
In addition to the first party Microsoft Copilot Agents – other ISVs and major software manufacturers are building a universe of 3rd party Copilot Agents. These will also hook into their own products, enabling Microsoft 365 Copilot to truly act as the “UI for your AI” in practice.
Finally, if you need to connect to a smaller COTS software package, or a LOB app that was developed in-house, you’ll need to build a custom Copilot Agent, which you can do (or leverage a partner like Cognizant to do) via Copilot Studio. Copilot Studio will allow developers to build Copilot Agents, as well as end-users themselves – so a huge range of Copilot Agents is possible.
What will all this cost? The answer, of course, is it depends.
Microsoft’s first party Copilot Agents today cost $20 standalone (if you already purchased licenses of Microsoft 365 Copilot), or $50 when included with Microsoft 365 Copilot. It is certainly possible that many of the 3rd party Copilot Agents built by ISVs may range in price from $0 up to $20 or more. For that cost, of course, you are getting Microsoft (or the ISV) to own the technical debt of keeping the agent up-to-date and functional with improvements to the product.
Of course, if you pay a partner like Cognizant to develop a Copilot Agent – you’ll also have to decide whether you want us to take on the technical debt for a cost, or if your own internal development team will take that on once the agent has been deployed. Many customers are in the early stages of using agents and as the new autonomous agents (see below) are deployed, in addition to technical debt, you will also need to understand the risk associated with developing your own powerful agents.
Why are Copilot Agents important?
Building on the foundational capabilities of Microsoft 365 Copilot, Copilot Agents are specialized AI entities designed to perform more complex and context-specific tasks. While Copilot provides general assistance across various applications, Copilot Agents are tailored to specific functions, offering a deeper level of interaction and support.
Enhanced Functionality
Copilot Agents bring a new layer of sophistication to the user experience. These agents can be programmed to handle intricate workflows, automate multi-step processes, and provide insights that are highly relevant to particular tasks. For instance, a Sales Copilot Agent can manage customer interactions, track sales metrics, and even suggest strategies based on historical data. Meanwhile, a Marketing Copilot Agent might assist in campaign management, content creation, and audience engagement.
Seamless Integration
One of the standout features of Copilot Agents is their seamless integration with Microsoft 365 Copilot. Users can interact with these agents within the familiar interfaces of their favorite Microsoft applications, ensuring a smooth and intuitive experience. This integration allows for a unified “UI for your AI,” where users can effortlessly switch between general and specialized AI assistance without disrupting their workflow. The biggest concern we have heard from customers who have made large investments in generative AI tools is that their staff are unwilling to go to a new webpage or interact with a new bot for each service. By standardizing on using Business Chat as the single interface for all AI interactions it makes everyone’s life much easier.
Personalized Assistance
Copilot Agents are designed to learn from user interactions, continually improving their performance and adapting to individual preferences. This personalized approach means that over time, the agents become more attuned to the specific needs and working styles of their users, providing increasingly relevant and effective support.
The Impact on Productivity
The introduction of Copilot Agents is set to transform productivity in profound ways. By offloading complex tasks to specialized AI agents, users can focus on higher-level strategic thinking and creative endeavors. This shift not only enhances individual productivity but also drives organizational efficiency and innovation.
Real-World Applications
Consider a project manager overseeing a large-scale initiative. With the help of a Project Management Copilot Agent, they can automate scheduling, resource allocation, and risk assessment. The agent can provide real-time updates, flag potential issues, and even suggest corrective actions, allowing the project manager to concentrate on overall strategy and stakeholder communication.
In the finance sector, a Financial Analysis Copilot Agent can streamline data analysis, generate comprehensive reports, and offer predictive insights. This level of assistance enables financial analysts to make data-driven decisions with greater speed and accuracy.
The business impact of autonomous agents
Microsoft previewed this week how Copilot Agents are evolving. The easiest way to think about agents is to classify them into three buckets:
The first two types of agents are already deployed at customers – with the third group coming in the next few weeks. Autonomous agents are to a large degree fulfilling the promise that many AI evangelists have made for years – truly able to start to drive businesses more quickly and more effectively to impact measurable business KPIs.
And that brings us to the final point I made in an earlier newsletter around Viva Insights – these Copilot Agents are going to change all the businesses they are deployed in – but in order to see that improvement, you’re going to need to stand up tools such as Viva Insights, tied to measurable business KPIs you defined beforehand, to see the benefits.
If you aren’t measuring these KPIs systematically yet, you’ll need to start down that road immediately. Haven’t defined the KPIs yet? We can help – at Cognizant our dedicated Consulting team has helped numerous clients define and track these business outcomes before you radically speed up with Copilot Agents.