Introducing Microsoft 365 Copilot
Copilot for Microsoft 365 represents a new paradigm in work, where AI and employees collaborate for increased productivity. It addresses the modern challenge where the pace of work is overtaking our ability to keep up. Statistics show that 64% of people struggle with time and energy for their jobs, and 70% would delegate work to AI to reduce workloads. Business leaders are increasingly recognizing the value of AI in boosting productivity rather than just reducing headcount.
Copilot for Microsoft 365 is designed to transform how we work in the digital age. Powered by Large Language Models (LLMs) and utilizing your business data from Microsoft Graph, it aims to spark creativity, boost productivity, and foster new skills. It integrates seamlessly with the Microsoft 365 suite, including Teams, Word, Outlook, PowerPoint, and Excel.
Examples of Copilot for Microsoft 365 in action may include:
- Outlook. Summarize the content of a large email thread.
- PowerPoint. Turn text-heavy slide into concise bullet points.
- Word. Rewrite a paragraph in a different tone or style.
* New Outlook for Windows For seamless integration of Copilot with Outlook, you’re required to use the new Outlook for Windows, currently in preview.
The Copilot System: Explained by Microsoft
Copilot’s understanding context and user needs
Copilot for Microsoft 365’s effectiveness stems from its unparalleled ability to understand you — the user. It does so by:
- Analyzing content. Whether it’s the document you’re drafting, the email you’re composing, or the meeting you’re in, Copilot scrutinizes the subject matter, tone, structure, and semantics to determine your intent and meaning.
- Gaining personal insights. Your profile information, communication patterns, and activity history help Copilot to understand your interests, expertise, and preferences.
- Real-time monitoring. As you work, Copilot continuously observes to gauge your immediate needs and discern how it can best assist you.
How Does It Work?
How does Copilot actually generate it’s responses based on your company data?
- A user provides an input or asks a question in an app, like Word or PowerPoint.
- Copilot prepares this input using a method known as “grounding.” This step ensures that responses aren’t general, but specific and relevant to your task. During this phase, Copilot consults Microsoft Graph to gather related data from your organization. Importantly, it only accesses data the user is already permitted to see based on their role and permissions in Microsoft 365.
- This data-gathering step is called “retrieval-augmented generation.” It’s how Copilot combines the user’s data with other relevant sources, like knowledge base articles, to refine the question for a more accurate answer.
- With the improved input, Copilot consults the Large Language Model (LLM) to generate an initial answer.
- Copilot further refines the answer before sending it back. This step involves more checks against Microsoft Graph, evaluations for responsible AI practices, security assessments, compliance checks, and even converting the answer into actionable commands.
- Finally, Copilot offers a well-formed, relevant suggestion back to the user, along with any actionable commands. The results you see are closely tied to your organization’s specific data, ensuring relevance and context.
Explore the core components of Copilot for Microsoft 365
By exploring some of the foundational elements that help power Copilot for Microsoft 365 we can gain a clearer understanding of the intricate processes that enable Copilot to offer its recommendations and suggestions.
Large Language Models
Large Language Models (LLMs) represent a class of artificial intelligence models that specialize in understanding and generating human-like text. The “large” in LLM signifies both the size of the models in terms of the number of parameters they encompass, and the vast volume of data on which they’re trained. LLMs, including models like ChatGPT, are a type of generative AI. Instead of merely predicting or classifying, generative AI can produce entirely new content. When applied to text, LLMs can generate contextually relevant and syntactically correct responses based on the provided prompts.
In the context of Copilot for Microsoft 365, LLMs are the engine that drivesit’s capabilities. Microsoft’s Azure OpenAI Service privately hosts these models, which Copilot for Microsoft 365 uses to understand user inputs and generate relevant responses.
Microsoft 365 keeps your data logically isolated by tenant (i.e. your company M365 account).
Natural Language Processing
Natural Language Processing (NLP) is a pivotal AI technology that helps machines understand, interpret, and respond to human language in a way that’s meaningful. In essence, NLP is the technology behind Copilot’s ability to read, comprehend, and generate text similar to how humans would. Some of the components involved are:
- Tokenization. Simplifies complex paragraphs by breaking down text into smaller chunks, like words or phrases.
- Semantic Analysis. Helps Copilot understand the underlying meaning or context.
- Sentiment Analysis. Assess the mood or emotion behind a text, Copilot can understand user intent more accurately.
- Language Translation. Aids in multilingual tasks, allowing Copilot to assist users across different languages.
NLP bridges the gap between human language and machine understanding. This technology ensures that when you ask Copilot something, it understands and responds effectively.
Semantic Index for Copilot
The Semantic Index for Copilot constructs an intricate map of your personal and company data, establishes important connections and identifies significant relationships. This design is much like the inner workings of the human brain. It goes beyond the confines of keyword search by interpreting and encoding the conceptual relationships between data elements. By analyzing your Microsoft Graph data – emails, documents, calendars, chats, etc. – and working with LLMs, it delivers more personalized, relevant, and actionable responses.
Microsoft Graph
Microsoft Graph is the connective tissue that binds all your Microsoft 365 services and data together. It’s a unified API that provides access to data and intelligence within Microsoft 365 from services like Outlook, OneDrive, SharePoint and Teams.
Copilot applies Microsoft Graph and the Semantic Index to synthesize and search content from multiple sources within your tenant. This means you don’t need to navigate away or switch apps – Copilot brings the relevant information to you.
Copilot takes user permissions, data security, and compliance seriously. It only generates responses based on the information you’re allowed to access. It achieves this goal through Microsoft Graph, a robust framework for secure data access, so is powerful capabilities align with your company’s security and privacy policies.
Summary
The above is (mostly) from Microsoft documentation and provides a useful overview of how Copilot for Microsoft 365 works, but we’ll have to start using it before we really know what to make of it and how far it can go to make our working lives better. I suspect it’s going to require a significant shift in our approach to how we work to really make the best of it, but time will tell.
There hasn’t been such a massive shift in gear in IT for a long time, possibly not since the early 2000’s when Microsoft’s Small Business Server brought the kind of in-house IT to smaller businesses that were previously only available to large corporates because the complexity and cost. Copilot, and others like it, could do the same for vast data processing.
We’re still in early days. For (a bit) more info, here’s the official Microsoft Blog from 15th Jan 2024. Bringing the full power of Copilot to more people and businesses – The Official Microsoft Blog
Explore how to use Copilot: Microsoft Copilot for Microsoft 365—Features and Plans
Try the Microsoft Learn Introduction for yourself: Introduction to Copilot for Microsoft 365 – Training | Microsoft Learn
See our other blog: Copilot and AI in Windows 11 – Macnamara ICT
