Reinvent the Modern Workplace With Microsoft 365 Copilot

  • June 15, 2023

How generative AI is evolving

Generative AI has gained significant attention in recent months and several tech giants are striving to become leaders in this space. ChatGPT reached one million users in five days due to its easy accessibility and natural language capability. Other popular platforms such as Twitter, Facebook, Spotify and Instagram took anywhere from two months to two years to achieve such a user base.

In AI, data becomes more valuable as it gets older. Primarily, it works on two models: the Foundation Model (FM) and the Large Language Model (LLM). The Foundation Model is a large machine learning model trained on massive amounts of unstructured and unlabeled data. Whereas the LLM is a specific foundation model trained on natural language. Applications like ChatGPT are based on the LLM model. Microsoft has also developed Microsoft 365 Copilot based on generative AI, which can reduce laborious tasks.

Microsoft 365 Copilot: a true intelligent assistant

In March 2023, Microsoft announced Microsoft 365 Copilot, a powerful everyday AI service combining the power of LLMs and enterprise data from Microsoft 365 ecosystem via Microsoft Graph. In May 2023, Microsoft launched an invite-based early access program of M365 Copilot.

Microsoft is integrating Copilot into a wide range of its products, including Dynamics 365 and popular productivity apps like Outlook, Excel, Word, PowerPoint, Loop and more. M365 Copilot is extending its capabilities to Power Automate and Power App to improve automation and efficiency. Copilot is also embedded into the experience management platform Viva and security solutions, similar to Microsoft Defender.

As an intelligent digital assistant, M365 Copilot can write email, take meeting notes, generate press releases, write code, analyze data in Excel, draft enterprise policies and create PowerPoint presentations.

Let’s consider the complex conceptual architecture of M365 Copilot, using a common task example:

User interface: Users can give natural language commands to Microsoft productivity apps like PowerPoint, Word, Excel, Outlook, Teams and others. For example, a user can say, “Create a proposal presentation for a client.”

Prompt engineering: M365 Copilot performs prompt engineering on inputs. It improves the prompt by retrieving business content and the context of a specific client through Microsoft Graph.

Contextualization: M365 Copilot sends the modified prompt to LLM to contextualize the request for the user.

Responsible AI check: M365 Copilot retrieves the response from LLM, makes a call to Microsoft Graph, performs a responsible AI check, conducts a security/privacy review and generates a command. In this case, to generate a PowerPoint presentation in the form of a proposal.

Output: M365 Copilot sends a summarized response to the user and gives commands to PowerPoint to generate a presentation based on the user’s content and client-specific context through Microsoft Graph.

Potential use cases of Microsoft 365 Copilot

Here are some key use cases from workplaces and customer experience management:

Productivity: M365 Copilot can significantly enhance the productivity of frontline workers, knowledge workers and contractors. Copilot for Viva Sales will streamline the sales process by providing deal summaries, drafting contextual email and granting control over writing style and length. Integration of M365 Copilot with Dynamics 365 will enable automated updates in the CRM system, allowing sales teams to invest more time on core activities.

Superior experience: M365 Copilot has the potential to revolutionize the meeting experience in the hybrid work model, promoting efficient decision-making and seamless collaboration. It summarizes missed Teams meetings, clarifies questions and crafts follow-up agendas based on chat history. The tool strategically highlights essential meeting content, actions, owners and client references, allowing participants to stay up to date.

Innovation: M365 Copilot democratizes low-code development by enabling individuals with any coding skill level to build applications, chatbots and complex workflows in Power Automate. Users can drive rapid innovation by swiftly creating, enhancing and automating routine tasks with the help of natural language descriptions. By reducing reliance on traditional coding, M365 Copilot accelerates the delivery of customized solutions, fosters agility and makes application development flexible.

Knowledge management: If integrated with a knowledge management system, M365 Copilot can answer frequently asked questions regarding employee benefits and leave policies in natural language without having to browse through multiple articles.

Personal coaching: Copilot in Microsoft Viva gathers upskilling requirements for assigned roles, creates customized learning plans and blocks calendars. It can help leaders streamline learning for their teams. Viva Topics provides employees with a conversational user interface to explore different topics.

Recommendations for getting started with Microsoft 365 Copilot

Planning and assessment: Start with one of the potential use cases described above. Conduct a comprehensive assessment of current Microsoft 365 landscapes, define measurable objectives and involve all the key stakeholders.

Pilot deployment: Undertake a pilot deployment in a controlled environment to test the Microsoft 365 Copilot functionality. Evaluate it against the objectives defined in the planning and assessment phase.

Security and compliance: Work with enterprise security team to review Microsoft 365 security settings in line with the enterprise security guidelines. Involve the security team at an early stage of the pilot deployment phase.

User training and change management: Develop a robust training plan and educate employees about the workplace experience that M365 Copilot can offer. Build a knowledge base to capture the lessons learned and keep updating it to avoid repeating issues. Establish a communication strategy to address users’ concerns/suggestions, manage resistance and drive adoption. Foster a culture of collaboration with the pilot team and encourage them to provide feedback.

Expansion: Take the deployment of M365 Copilot to a larger user group with more use cases after the knowledge base has been updated with lessons learned in pilot deployments.

While keeping an eye on the future of generative AI, NTT DATA is incorporating additional products into its suite of digital workplace products that are built upon its AI-enabled Nucleus platform. We aim to elevate user experience and increase productivity in modern workplaces.

Get access to your Microsoft 365 Copilot advisory Workshop, designed to unleash true business value with generative AI technology or contact us to modernize your workplace.


Subscribe to our blog

Sandeep Kumar
Sandeep is the Digital Workplace Strategic Advisor, focused on driving productivity in hybrid work and lowering the TCO by integrating and automating the digital workplace ecosystem. He helps organizations make data-driven decisions, delivering operational insights by harnessing the data using a cloud analytics engine powered by AI and ML.

Related Blog Posts