Support the Podcast
Contacts
Info
M365.FM is a podcast about Microsoft 365, Microsoft Copilot, AI, Modern Work, security, governance, Power Platform, Azure, and the technologies shaping the future of work.Hosted by Microsoft MVP Mirko Peters,...
show more
M365.FM is a podcast about Microsoft 365, Microsoft Copilot, AI, Modern Work, security, governance, Power Platform, Azure, and the technologies shaping the future of work.Hosted by Microsoft MVP Mirko Peters, M365.FM brings together Microsoft MVPs, Microsoft employees, product experts, architects, developers, and community leaders from around the world.Each episode goes beyond announcements and hype to explore what Microsoft technologies mean in practice. From Microsoft 365 Copilot and AI agents to Teams, SharePoint, Power Platform, Microsoft Fabric, Entra, Purview, security, governance, adoption, and automation, M365.FM focuses on real-world experience, implementation, strategy, and lessons learned.Expect expert interviews, technical deep dives, practical explainers, and conversations with people building, implementing, and shaping the Microsoft ecosystem.If you work with Microsoft 365, Copilot, AI, Modern Work, or the Microsoft Cloud, M365.FM helps you understand what matters, what works, and what is coming next.Hosted by Mirko Peters, Microsoft MVP.
show less
M365.FM a Microsoft MVP Podcast by Mirko Peters
M365.FM a Microsoft MVP Podcast by Mirko Peters
Mirko Peters (M365 Consultant)Support the Podcast
Episodes & Posts
Episodes
Posts
Transcribed
3 SEP 2026 · Artificial intelligence is changing software development at a speed we have rarely seen before.Developers have already moved from writing every line of code themselves to working with AI assistants that can generate code, explain unfamiliar systems, create tests, debug applications, and automate repetitive work.But according to Microsoft MVP Andre Baltieri, that is only the beginning.In this episode of M365.FM, Mirko Peters sits down with Andre for a deep dive into the transition from traditional software development to AI-assisted development, coding agents, agentic architectures, Microsoft Agent Framework, .NET, RAG, context engineering, security, and the economics of generative AI.
FROM .NET IN 2003 TO THE AI ERA
Andre takes us back to the early days of .NET and C#, when learning a new Microsoft technology often meant purchasing official training and traveling to another city.Since then, software development has moved through desktop, web, mobile, cloud, containers, microservices, and serverless computing.Andre argues that the transition to AI feels fundamentally different. Instead of simply introducing another platform or framework, AI introduces a new way for humans to interact with software.
AI ASSISTANTS VS. AI CODING AGENTS
There is an important difference between having an AI assistant inside your IDE and delegating work to an agent.An assistant can explain code, suggest refactoring, answer questions, and help developers understand their applications.An agent can receive a goal, create a plan, divide the work into smaller tasks, use tools, coordinate additional agents, and implement significant parts of the solution.Andre explains how this is already changing his own development workflow, with AI now generating much of the code he previously would have written manually.
SPEC-DRIVEN SOFTWARE DEVELOPMENT
As agents become more capable, specifications become increasingly important.Instead of describing every implementation detail, developers can define requirements, architecture, constraints, and expected behavior and allow agents to determine how parts of the implementation should be completed.This shifts developer attention from simply producing code toward defining what should be built and why.
MICROSOFT AGENT FRAMEWORK
The conversation moves into Microsoft Agent Framework and its role in bringing AI capabilities into existing applications.Andre explains how the framework brings together capabilities associated with Semantic Kernel and AutoGen and provides developers with tools for connecting models, orchestrating workflows, using MCP, implementing RAG, handling data ingestion, and exposing application functionality to AI.For .NET developers in particular, this can significantly reduce the amount of integration code required.
WHY .NET STILL MATTERS IN THE AI ERA
Python remains one of the dominant languages in AI development, but Andre argues strongly that .NET and C# are extremely well positioned for enterprise AI applications..NET continues to evolve rapidly, while Microsoft's AI tooling increasingly gives C# developers native access to modern AI capabilities.Organizations with years of business logic already implemented in .NET may therefore have a major advantage: they do not necessarily need to rebuild everything before introducing AI.Existing functionality can instead be selectively exposed to agents and AI-powered applications.
FROM DETERMINISTIC SOFTWARE TO AGENTIC SYSTEMS
Traditional applications are largely deterministic:If X happens, execute Y.Agentic systems introduce another model:Here is the goal. Determine which actions are required to accomplish it.That represents a significant architectural shift.Instead of explicitly defining every possible path, developers increasingly define goals, tools, context, permissions, constraints, and boundaries within which AI can operate.
DESIGN PATTERNS ARE NOT DEAD
AI-generated code does not eliminate decades of software engineering knowledge.Clean code, maintainability, testing, architecture, and design patterns remain important because AI frequently learns how to implement new functionality by examining the existing codebase.Messy code can therefore lead to more messy code.Developers still need to understand architecture and engineering principles even when an AI agent performs much of the implementation.
THE STOCHASTIC SOFTWARE PROBLEM
Traditional developers expect identical inputs to produce identical outputs.Generative AI is probabilistic.The same request can produce different implementations, answers, or behavior across multiple executions.Andre discusses why this requires developers to rethink testing and validation and why strong guardrails become increasingly important when AI functionality is exposed to large numbers of users.
CONTEXT ENGINEERING IS MORE IMPORTANT THAN PROMPTING
Choosing the latest model is not necessarily the most important decision.Andre argues that context is everything.Developers need to understand both the business problem and the technical environment well enough to provide AI with the right information.Too little context produces weak results.Too much context can overwhelm the model.The challenge is finding the information that actually matters.
RAG, DATA AND THE CONTEXT WINDOW
Retrieval-Augmented Generation becomes especially important when organizations want AI systems to work with their own knowledge.But building a RAG system is not simply about putting documents into a vector database.Data needs to be cleaned, structured, chunked, retrieved, and inserted into the model's context intelligently.Andre shares an example from his own education platform, where video lessons were transcribed and indexed so users could search for concepts and jump directly to the relevant point in a video.
MEMORY AND MANAGING AI CONTEXT
Long-running AI conversations create another challenge: memory.As context windows fill, conversations need to be summarized or compacted.Andre explains why developers should actively manage this process instead of assuming that an AI system will always preserve the most important information.Sometimes the best solution is surprisingly simple: finish a task, close the conversation, and start again with a clean context.Specifications and Markdown files can also provide persistent project context for coding agents.
SECURITY, PERMISSIONS AND LEAST PRIVILEGE
Giving an AI agent access to tools and company data creates significant security implications.Andre recommends treating agents according to principles similar to human identities: close everything by default and expose only what the agent genuinely requires.Instead of giving an AI system unrestricted database access, developers should expose carefully controlled functions that return only the information required for a particular task.This becomes particularly important when agents can read or modify enterprise data.
PROMPT INJECTION AND AI GUARDRAILS
Prompt injection creates a new attack surface for AI-powered applications.Users can intentionally manipulate prompts, attempt to retrieve information outside the intended context, consume company resources, or persuade an AI system to perform actions its designers never anticipated.The discussion explores the importance of system instructions, application-level restrictions, controlled functions, identity, permissions, and platforms such as Azure AI Foundry for establishing additional security boundaries.
AI FINOPS — DON'T USE GENERATIVE AI FOR EVERYTHING
One of the most practical lessons from the conversation is that just because AI can perform a task does not mean AI should perform that task.Andre distinguishes between generative and deterministic workloads.If something must happen the same way every time, traditional programming may be faster, cheaper, and more reliable.He gives the example of his video workflow: Python scripts can extract audio and perform deterministic processing locally, while generative AI is reserved for tasks such as translation where generation actually adds value.The result is a hybrid architecture that can dramatically reduce unnecessary token consumption.
BUILDING THE AGENTIC FUTURE
Software development is moving beyond developers manually defining every individual step.Increasingly, developers will define goals, specifications, context, tools, permissions, architecture, and guardrails while AI systems determine how portions of the work should be accomplished.That does not eliminate the developer.It changes where the developer creates value.Understanding the business, designing maintainable systems, controlling context, securing tools and data, validating AI-generated work, and deciding when not to use AI may become some of the most important software engineering skills of the agentic era.RAPID FIRESingle agent or multi-agent?For complex workloads, Andre sees significant potential in multi-agent architectures and sub-agents.Prompt engineering or context engineering?Context engineering.And what comes next?More capable models, more powerful agents, better code generation, stronger architectures, and continued evolution of the tools developers use to build software.We are still at the beginning of the generative AI era.
ABOUT THE GUEST
Andre Baltieri is a Microsoft MVP and software development specialist with more than two decades of experience in the industry.His work focuses on .NET, C#, artificial intelligence, Microsoft Agent Framework, software architecture, and modern AI-assisted development.In this conversation, he brings together more than twenty years of software engineering experience with a practical view of how AI agents are changing the developer profession. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Transcribed
2 SEP 2026 · Microsoft Power Platform is often described as a low-code platform. But what happens when the applications you build become business-critical, highly integrated, and too complex for a simple maker-first approach?In this episode of M365 FM, Mirko Peters talks with Power Platform Solution Architect Ian Tweedie about what happens when Power Platform moves beyond simple low-code applications and becomes part of a serious enterprise architecture.
LOW-CODE DOESN’T MEAN LOW ARCHITECTURE
Power Platform can deliver a large percentage of business value quickly, but enterprise solutions almost always contain requirements that go beyond standard low-code capabilities.Ian explains why low-code should never be confused with no-code — and why traditional software architecture principles still matter when building with Power Apps, Power Automate, Dataverse, custom connectors, APIs, and Azure services.
WHEN POWER PLATFORM BECOMES ENTERPRISE SOFTWARE
There isn’t necessarily a clean line between “low-code” and “enterprise.”Complexity starts increasing when applications involve multiple user journeys, development teams, integrations, security requirements, business-critical processes, and interconnected services.At that point, architecture becomes essential.Ian explains why solutions should be divided into clearly defined features and modules with clean interfaces instead of becoming one large interconnected application.
ESCAPING THE WHACK-A-MOLE DEVELOPMENT PROBLEM
Fix one bug and another appears somewhere else.That familiar development problem is often a symptom of tightly coupled architecture.Ian discusses how modular design can isolate functionality and reduce unintended dependencies. Using email delivery as an example, he explains how separating business processes from delivery mechanisms can make applications easier to test, maintain, replace, and scale.
LOW-CODE + PRO-CODE = HYBRID ARCHITECTURE
Power Platform doesn’t have to compete with traditional software development.A Power Apps frontend might represent only a small part of a much larger application using Azure, AWS, GCP, APIs, or custom services.The important architectural question isn’t whether something is “low-code” or “pro-code.”It’s which technology is best suited to each feature.
CITIZEN DEVELOPERS, MAKERS AND ARCHITECTURE
Citizen developers bring something extremely valuable: deep knowledge of the business processes they work with every day.But business expertise doesn’t automatically translate into good application architecture.Ian discusses the balance between empowering makers and introducing enough architecture, governance, normalization, and technical review to prevent solutions from becoming difficult to maintain.
GOVERNANCE WITHOUT KILLING INNOVATION
Too little governance creates chaos.Too much governance creates friction — and can drive employees toward unsupported workarounds, spreadsheets, VBA, and shadow IT.The challenge is finding the right level of governance based on organizational risk, application criticality, users, and business impact.
POWER PLATFORM, APIs AND AZURE
Where should business logic live?Should secrets be stored inside Power Platform? When should Azure Key Vault, Azure Functions, or API Management become part of the architecture?Ian explains why architecture should always begin with the problem being solved rather than adding Azure services simply because they are available.
GIT, SOURCE CONTROL AND CI/CD
As Power Platform development becomes more collaborative, traditional development practices become increasingly relevant.The conversation explores Git, repositories, development environments, pipelines, source control, feature isolation, cross-dependencies, and CI/CD.There may not always be a perfect approach to source control in Power Platform — sometimes the goal is choosing the “least worst option” for the project.
IAN’S POWER PLATFORM ARCHITECTURE PLAYBOOK
Ian’s core principle is straightforward:Break solutions into features and modules.Each feature should have a clear reason to exist and ideally perform one specific responsibility.Then determine how those features communicate, where logic should execute, how they should be tested, and which technology is best suited to implementing them.
THE RAPID-FIRE ROUND
Mirko puts Ian through a series of quick questions:Should every enterprise application use Dataverse?Can Power Platform build mission-critical applications?Can Power Automate replace Logic Apps?Does low-code automatically reduce technical debt?Can Power Platform replace traditional application development?And perhaps most importantly: what should you order when visiting Newcastle?
KEY TAKEAWAY
Low-code does not mean low architecture.As Power Platform solutions become larger, more connected, and more important to the business, the fundamentals of software engineering remain relevant.Architecture matters.Data modeling matters.Governance matters.Security matters.API design matters.DevOps matters.And successful enterprise Power Platform development is increasingly about understanding how low-code, pro-code, Azure, APIs, automation, and traditional software engineering practices work together. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Transcribed
1 SEP 2026 · Microsoft 365 Copilot doesn’t arrive in a clean Microsoft 365 tenant. It arrives in an environment that organizations have been building for years — full of SharePoint sites, Teams, OneDrive content, duplicate documents, historic migrations, guest accounts, inconsistent permissions, forgotten workspaces, broken ownership models, and content nobody has reviewed in years.For traditional Microsoft 365 management, much of this technical debt could remain relatively hidden. AI changes that.In this episode of M365.FM, Mirko Peters talks with Richard Harbridge, Microsoft MVP and industry advisor at ShareGate, about why AI readiness is fundamentally connected to Microsoft 365 modernization and governance.The conversation moves beyond the question of how to deploy Copilot and focuses instead on a more important question: What kind of Microsoft 365 environment are we actually giving AI access to?
COPILOT DOESN’T CREATE THE MESS — IT AMPLIFIES IT
One of the central ideas in this conversation is that Microsoft Copilot is not necessarily creating entirely new governance problems. Instead, AI makes existing problems significantly more visible and consequential.Organizations have accumulated years of decisions around permissions, sharing, workspace creation, ownership, migrations, external access, inactive content and collaboration.Richard describes this accumulation using the concept of “sprawl.”Sprawl itself is not automatically bad. A growing number of Teams, SharePoint sites and other resources can be evidence that people are successfully adopting Microsoft 365.The problem begins when that growth happens faster than the organization’s ability to manage it.Temporary permissions become permanent. Old collaboration spaces remain accessible. Ownership changes. Business structures evolve. Content stays online long after its original purpose has disappeared.Copilot then operates on top of that existing environment.
MICROSOFT 365 SPRAWL IS BIGGER THAN TEAMS AND SHAREPOINT
When people hear “Microsoft 365 sprawl,” they often immediately think about too many Teams or SharePoint sites.But Richard argues that sprawl is multidimensional.Organizations can experience workspace sprawl, permission and access sprawl, ownership problems, inactive resources, lifecycle problems, administrative complexity, conditional access sprawl and increasingly integration and connector sprawl.The rise of AI adds another dimension.Microsoft 365 environments increasingly connect data, applications, AI systems and agents. That means organizations need to think beyond individual workloads and start looking at Microsoft 365 as an interconnected ecosystem.Governance can no longer be treated purely as “SharePoint governance” or “Teams governance.”
THE CONFIDENCE GAP IN MICROSOFT 365 GOVERNANCE
Richard shares an especially interesting finding from ShareGate’s research.A very large percentage of IT leaders report being highly confident in their Microsoft 365 governance. Yet a significant portion of those same organizations report that Copilot has surfaced content that users arguably should not have been able to discover — or they suspect this may have happened but do not know how to verify it.That creates an important distinction between having governance controls available and actually having a continuously governed environment.Policies, configuration options and administrative controls alone do not guarantee that an organization understands its current Microsoft 365 state.AI can expose that gap very quickly.
DID MICROSOFT MAKE COLLABORATION TOO EASY?
Microsoft Teams, SharePoint and OneDrive are successful partly because Microsoft has made collaboration extremely easy.But easy collaboration also makes it easy to create more resources.One response organizations have traditionally used is restricting workspace creation. Richard explains why that alone does not solve the problem.A workspace created today may serve an entirely different purpose six months, one year or two years later.Its owners may change.The organization may restructure.The project may finish.The content may become irrelevant.The people who originally understood why the workspace existed may leave.Governance therefore cannot stop at provisioning.Organizations need lifecycle processes, reviews, attestations and signals that continuously determine whether resources are still appropriate.
GOVERNANCE HAS TO BECOME CONTINUOUS
One-time cleanup projects are not enough.An organization can spend months cleaning its Microsoft 365 environment before deploying Copilot, but without continuous governance, the same problems will gradually return.That means organizations need repeatable processes around ownership, permissions, inactivity, lifecycle management, archiving and retirement.The objective is not to eliminate Microsoft 365 sprawl completely.It is to turn unmanaged sprawl into managed sprawl.
WHERE SHOULD ORGANIZATIONS START BEFORE SCALING COPILOT?
Imagine an enterprise with 10,000 employees, thousands of Teams and SharePoint sites, years of accumulated content and inconsistent permissions.Where should it begin?Richard recommends looking at the specific types of sprawl and identifying where the organization carries the greatest risk.For one company, that might be access control and oversharing.For another, it might be inactive content and lifecycle management.For another organization, privileged identities or administrative management may represent the larger problem.Instead of treating Microsoft 365 governance as one enormous cleanup exercise, organizations can break the problem into measurable categories and prioritize the areas where improvement matters most.
COPILOT ADOPTION SHOULD BE TEAM-BASED
The conversation also challenges one traditional Microsoft technology adoption model.Organizations frequently deploy new technologies through distributed champions.Richard argues that Copilot benefits from a more team-oriented adoption model.Rather than distributing a small number of licenses across unrelated champions throughout an organization, companies can benefit from saturating teams with AI capabilities so that people learn together, develop shared practices and integrate Copilot into collaborative workflows.AI adoption is not only about giving individuals another productivity tool.It changes how teams work together.
MAKING MICROSOFT 365 RISK MEASURABLE
Governance initiatives often struggle because their value is difficult to communicate to business leaders.The conversation explores ShareGate’s Risk Radar, which is designed to help organizations evaluate different categories of Microsoft 365 sprawl, compare maturity against benchmarks and understand which areas deserve attention.An important part of this approach is translating governance risk into financial terms.Instead of asking leadership for time and resources simply because “we need better governance,” IT teams can connect improvements to the potential cost of unmanaged risk.That makes Microsoft 365 modernization and governance easier to position as a business investment rather than another administrative IT project.
AI AGENTS MAKE GOVERNANCE EVEN MORE IMPORTANT
Copilot is only part of the story.As organizations begin creating and deploying more AI agents, governance becomes substantially more complex.Agents can change over time. They can interact with tools, data and other systems. Their ownership can become unclear. Permissions can evolve, and organizations will eventually have to deal with agent lifecycle management at scale.Who owns an agent?Who reviews it?What resources can it access?What happens when its original owner leaves?When should an agent be retired?When should multiple agents be consolidated?These questions look remarkably similar to problems organizations already experience with Teams, SharePoint sites and other Microsoft 365 resources — except AI increases both the speed and potential impact.
GOVERNANCE FROM THE RESOURCE UP
Richard discusses the importance of looking at governance from the underlying resources upward.An AI experience may sit on top of SharePoint, Microsoft 365 data, Power Platform resources, connectors, permissions and other systems.Organizations therefore cannot govern only the visible AI interface.They need to understand the complete chain of resources, data and permissions supporting it.This becomes increasingly important as agents gain more capabilities and organizations move from relatively simple assistants toward agents that can take action.
THE FUTURE IS “EVERYTHING GOVERNANCE
”Looking several years ahead, Richard expects Microsoft governance to become significantly more holistic.Instead of separate conversations about SharePoint governance, Teams governance, Power Platform governance and AI governance, organizations increasingly need to understand how all of these layers interact.Microsoft Purview is highlighted as an important part of this evolution, particularly as Microsoft expands capabilities around data security and posture management.Governance professionals therefore need to broaden their perspective beyond the administration interface or workload they traditionally specialized in.AI sits across boundaries.Governance increasingly has to do the same. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Transcribed
30 AUG 2026 · Does Microsoft 365 Copilot really make people more productive?In this episode of the M365 FM podcast, Mirko Peters speaks with Microsoft MVP Adrian Espes about what Copilot actually delivers in everyday work—not in a polished demo, but during a normal working day filled with meetings, emails, documents, Teams messages, deadlines, and constant interruptions.Adrian brings a practical perspective shaped by his experience in training, business analytics, modern workplace consulting, and Microsoft 365 adoption. Together, Mirko and Adrian look beyond the marketing and explore where Copilot is genuinely useful, where expectations are still unrealistic, and what organizations need to consider before rolling it out.
MICROSOFT 365 COPILOT IN THE REAL WORLD
Adrian explains why Microsoft 365 Copilot’s biggest advantage is its position inside the Microsoft ecosystem. Unlike standalone AI tools, Copilot can work directly with the applications people already use every day, including Outlook, Teams, Word, PowerPoint, Excel, and Edge.The conversation explores how Copilot can help users summarize meetings, catch up on conversations, identify follow-up tasks, work with documents, and create useful outputs without constantly moving information between different tools.
PROMPTING WITHOUT BECOMING A PROMPT ENGINEER
Do employees really need to become prompt engineers?Adrian shares a realistic approach to prompting. Users do not need to learn complicated formulas or memorize a perfect prompt structure. Instead, they should practice explaining their goal clearly, provide relevant context, describe the desired outcome, and refine their requests over time.The discussion also covers why different departments and roles may need different prompting approaches. A financial analyst, a marketing professional, a manager, and a frontline worker will all use Copilot differently because their goals and daily tasks are different.
PRODUCTIVITY, EFFICIENCY, AND THE HUMAN FACTOR
One of the central themes of this episode is the difference between productivity and efficiency.Adrian explains why the word “productivity” can create anxiety among employees. When companies talk about productivity, many workers may fear that AI is being introduced to measure performance or replace jobs.Instead, Adrian suggests focusing on efficiency, better work quality, reduced friction, and making AI a natural part of everyday work. The real question is not simply whether someone completes more tasks, but whether Copilot helps them work with less effort, better information, and more time for meaningful activities.
DATA QUALITY, GOVERNANCE, AND SECURITY
Copilot can only be as useful as the information available to it. That makes data quality, permissions, governance, and information architecture essential parts of any Microsoft 365 Copilot project.Mirko and Adrian discuss the importance of reviewing the Microsoft 365 environment before implementation. This includes checking permissions, overshared information, tenant configuration, data protection, sensitivity labels, and the way users store and access content.Adrian also explains the importance of using enterprise accounts and understanding how enterprise data protection works when employees use Microsoft Copilot in a business environment.
MICROSOFT 365 COPILOT AND AI MODELS
The conversation also looks at the growing number of AI tools and models available today, including Microsoft Copilot, ChatGPT, Claude, Gemini, Perplexity, and different models available through GitHub and Microsoft platforms.Rather than asking which AI tool is universally the best, Adrian recommends choosing the right tool for the task. Some tools may be stronger for coding, reasoning, image creation, or creative work, while Microsoft 365 Copilot’s major strength is its integration with business data and workplace applications.
A PRACTICAL COPILOT IMPLEMENTATION ROADMAP
Buying Copilot licenses is only the beginning.Adrian outlines the considerations organizations should address when planning a Microsoft 365 Copilot rollout. Before investing in additional licenses, companies should first understand what is already available through their existing Microsoft 365 plans and evaluate whether users genuinely need the full Copilot experience.The implementation process should also include:Reviewing the Microsoft 365 tenant and existing configurationsChecking permissions and data governanceUnderstanding regulatory and regional requirementsEvaluating which users and roles will benefit mostDefining realistic use casesSupporting employees through training and experimentationMeasuring efficiency and adoption instead of relying only on license usage
WILL COPILOT ELIMINATE MEETINGS?
During the rapid-fire section, Mirko asks Adrian whether Copilot will eliminate most meetings.Adrian’s answer: for now, this is still mostly hype. Copilot can make meetings easier to follow, summarize discussions, and identify actions, but it does not automatically solve the organizational reasons why too many meetings exist.They also discuss whether AI agents will replace traditional business applications. Adrian believes agents will become increasingly important, but they will work alongside business applications rather than replace them entirely. Strong governance, security, visibility, and management will be essential as organizations create more agents.
IS COPILOT USEFUL FOR FRONTLINE WORKERS?
The value of Copilot for frontline workers depends heavily on their role and daily responsibilities.For employees who regularly work with email, Teams, documents, or operational information, Copilot may provide real benefits. However, not every frontline worker uses Microsoft 365 applications in the same way as an office-based employee.Adrian explains why organizations should avoid assuming that one Copilot strategy will work for every employee group. Adoption needs to be connected to real tasks, real users, and real business needs.
KEY QUESTIONS DISCUSSED IN THIS EPISODE
Does Microsoft 365 Copilot really improve productivity?What makes Microsoft 365 Copilot different from ChatGPT, Claude, Gemini, and other AI tools?Do employees need to become prompt engineers?How can users improve their prompts?Why do data quality and governance matter so much?How should organizations prepare for a Copilot rollout?What should companies evaluate during the first 30, 60, and 90 days?Can Copilot reduce meetings?Will AI agents replace traditional business applications?Is Copilot useful for frontline workers?How can organizations measure efficiency without creating fear among employees?
ABOUT ADRIAN ESPES
Adrian Espes is a Microsoft MVP focused on Microsoft 365 and Copilot. He works as a consultant and trainer, helping organizations understand, adopt, and use Microsoft technologies in practical business environments.His background includes sales, training, data analytics, business workflows, and modern workplace consulting. Adrian is passionate about helping people use AI more naturally and effectively in their daily work.
FINAL THOUGHTS
Microsoft 365 Copilot is not a magic productivity button. Its value depends on the quality of an organization’s data, the clarity of its use cases, the preparation of its environment, and the willingness of employees to experiment and learn.The most successful Copilot implementations will not focus only on buying licenses or showcasing impressive demos. They will focus on helping people work more efficiently, make better decisions, reduce repetitive effort, and use AI as a practical part of the modern workplace.Listen to this episode to discover what Microsoft 365 Copilot can really do beyond the hype.
ABOUT THE M365 FM PODCAST
The M365 FM podcast explores the people, ideas, and technologies shaping the future of work across Microsoft 365, Copilot, AI, security, governance, Power Platform, and the modern workplace.Hosted by Mirko Peters, every episode features conversations with Microsoft MVPs, product experts, consultants, architects, and practitioners from across the global Microsoft ecosystem.Subscribe to M365 FM for practical conversations about what works, what does not, and what organizations need to know next. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Transcribed
29 AUG 2026 · Microsoft 365 has given organizations more tools than ever to communicate, collaborate, automate, and manage information. Now Microsoft 365 Copilot adds an entirely new layer of AI capability. But behind all the discussion about productivity, automation, agents, and AI, there is a much bigger question: are these technologies simply helping us work faster, or are they fundamentally changing the way we think, learn, communicate, collaborate, and work?In this episode of m365.fm, Mirko Peters sits down with Tracy van der Schyff to explore the human side of Microsoft 365 and Copilot. Tracy has spent years working at the intersection of technology, productivity, digital literacy, training, adoption, and organizational change. Her focus is not simply on teaching people which buttons to click. It is about helping people understand technology, use it with purpose, and become more capable because of it.The conversation goes far beyond Copilot features. Mirko and Tracy discuss digital fluency, Microsoft 365 adoption, information management, AI readiness, change management, responsible AI, the broken digital workplace, and a powerful idea that runs throughout the episode: we design technology, but the technology we create and use also shapes us.
FROM DIGITAL LITERACY TO DIGITAL FLUENCY
For many years, organizations talked about PC literacy and later digital literacy. But Tracy believes those terms no longer fully describe the skills people need in a modern workplace.Knowing how to operate a computer is not enough. Knowing where a button is inside Microsoft Teams or how to upload a document to SharePoint does not necessarily mean somebody understands how to work effectively in a digital environment.Digital fluency goes further. It means understanding the purpose of a technology, knowing when it should be used, understanding how your actions affect other people, and using technology responsibly and intentionally.This becomes even more important as AI enters everyday work. Tracy explains how her original model of eight pillars of digital literacy has evolved into eleven pillars of digital fluency, now incorporating responsible AI use and the additional skills people need when working alongside AI systems.The important distinction is that these are not simply Microsoft 365 skills. They are becoming life skills.AI is also creating unexpected opportunities to develop human skills. Communicating effectively with Copilot requires people to explain what they actually want. Better questions can produce better answers, and those answers can help people formulate better questions the next time. In that sense, working with AI can improve communication, creativity, critical thinking, and the ability to express intent clearly.
THE MICROSOFT 365 TOOL OVERLOAD PROBLEM
Teams, Outlook, SharePoint, OneDrive, Loop, Planner, Lists, Power Platform, Copilot and countless other applications give employees enormous capabilities. But providing access to tools does not automatically teach people how those tools should fit together.Organizations often deploy technology and expect employees to figure out the rest.That can result in departmental information being shared from personal OneDrive accounts, Teams being created simply for individual meetings, documents being stored in inappropriate locations, and employees constantly switching between tools without understanding where work actually belongs.When that happens, Tracy argues that blaming users is the wrong response.If an employee was never taught the intent behind Teams, SharePoint, OneDrive, Outlook, or another application, they will naturally choose whatever tool helps them complete the immediate task. The underlying problem is often not the employee. It is the absence of a clear digital strategy.
ME, WE, US: SIMPLIFYING THE DIGITAL WORKPLACE
One framework Tracy uses to make the Microsoft 365 environment easier to understand is ME, WE, US.The ME space represents the individual. These are the tools and information primarily associated with your personal work.The WE space begins when people collaborate around a common goal. Microsoft Teams and collaborative SharePoint environments become particularly relevant here.The US space represents information intended for the broader organization, such as publishing environments and intranets.This sounds simple, and that is precisely the point.Employees should not need to understand every architectural detail behind Microsoft 365 before they can make a sensible decision about where their work belongs. Organizations need to translate complicated technology landscapes into models that ordinary employees can understand and apply.
COPILOT DOES NOT FIX A BROKEN DIGITAL WORKPLACE
One of the strongest messages from the conversation is that Microsoft 365 Copilot should not be treated as a solution for an already broken digital workplace.Copilot does not magically repair years of poor information architecture, excessive sharing, unmanaged Teams, confusing permissions, duplicated documents, abandoned SharePoint sites, or inconsistent working practices.Instead, Tracy describes AI as an amplifier.If an organization has a healthy digital environment, Copilot can amplify that healthy environment. If the underlying environment is unhealthy, AI can make those existing problems significantly more visible.In that sense, Copilot acts like a huge spotlight.Information that may previously have been difficult to discover can suddenly become much easier to surface. That can make organizations believe Copilot created a problem when, in reality, the underlying permissions, sharing practices, or information-management problems may have existed for years.The AI did not necessarily create the mess. It exposed it.
ONTOLOGICAL DESIGN: THE THINGS WE CREATE CHANGE
USA fascinating part of the conversation explores ontological design.The concept sounds complicated, but Tracy explains it in a very practical way: we create things, and the things we create eventually change us.That applies to software, applications, intranets, processes, social media, Microsoft Teams, Copilot, and almost every digital environment people interact with.When someone designs an application, they make decisions about how users will interact with it. Those decisions influence the behavior of the people using the application.The same principle applies at a much larger scale to the modern workplace.Employees can spend many hours every day inside Teams, Outlook, SharePoint, OneDrive, Microsoft 365, and increasingly Copilot. Those environments are therefore not neutral. They influence how people communicate, how quickly they expect responses, how they organize information, how they collaborate, and even how they think about work.The intention behind what we design matters because what people consume eventually influences them.
WHEN ACTIVITY BECOMES CONFUSED WITH PRODUCTIVITY
Microsoft Teams and other collaboration tools have dramatically reduced the friction required to communicate. But reducing friction can also create enormous amounts of noise.Tracy raises an interesting problem: people can begin to use visible activity as proof that they are productive.More messages. More notifications. More updates. More documents. More meetings. More channels.But every communication creates work for somebody else.A message that takes one person a few minutes to write might interrupt dozens or hundreds of other employees. Individually, that interruption may seem insignificant. Across an organization, the cumulative cost can become substantial.The question therefore should not simply be whether Microsoft 365 allows us to communicate faster. Organizations also need to ask whether all of that communication is necessary in the first place.
AI CAN BE MORE THAN A PRODUCTIVITY TOOL
Many discussions about Copilot focus on straightforward productivity scenarios.Summarize my emails. Summarize this meeting. Create a presentation. Rewrite this document. Analyze this information. Help me find something.Those capabilities are useful, but Tracy argues that AI can become something much more interesting.People can use AI to learn.Someone who struggles with delegation can ask Copilot how to become better at delegating. Someone intimidated by AI can ask AI how to begin learning about AI. Someone concerned about cybersecurity can ask how to protect themselves and their family. Someone who lacks confidence with a technology can use AI as a private environment for experimentation and learning.This moves the conversation from simply asking, “How much time can Copilot save?” toward asking, “How can Copilot increase human capability?”That distinction is important.The most valuable use of AI may not always be automating another task. Sometimes it may be helping someone become better at performing that task themselves.
AI DOES NOT REMOVE HUMAN RESPONSIBILITY
Copilot might summarize hundreds of emails, but responsibility does not disappear if an important message is missed.AI can generate a document quickly, but somebody still needs to understand why that document exists.AI can retrieve organizational information, but companies still need to understand permissions, ownership, governance, security, and information management.AI can produce an answer, but people still need the judgment required to evaluate that answer.For Tracy, this is another reason digital fluency becomes more important as AI becomes more capable.Organizations should not respond to increasingly powerful technology by investing less in human skills. They should invest more. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Transcribed
28 AUG 2026 · Modern SharePoint development does not mean building every component from scratch. In this episode of the M365 FM Podcast, Mirko Peters talks with Microsoft MVP Siddharth Vaghasia about building production-ready SharePoint Framework (SPFx) solutions by combining the right Microsoft 365 technologies with reusable community components.Siddharth brings nearly 18 years of experience across the Microsoft technology stack, from .NET and the early days of SharePoint Server to SharePoint Online, Microsoft 365, Power Platform, Azure, and modern SPFx development.
WHEN SHOULD YOU ACTUALLY USE SPFx?
Not every SharePoint requirement needs custom development. Siddharth explains a simple principle: first determine whether Microsoft already provides the functionality. If the requirement can reasonably be solved with standard SharePoint capabilities, avoid unnecessary customization.SPFx becomes valuable when organizations need experiences, integrations, or interfaces that cannot be delivered effectively with out-of-the-box functionality.
SPFx VS POWER PLATFORM
Should you build the solution with SPFx or Power Apps?The discussion explores where each approach fits. Power Apps can be effective for relatively straightforward forms, conditional fields, business rules, and scenarios where citizen development and low-code maintainability matter.SPFx becomes particularly powerful when developers need greater control over the user interface, complex data handling, reusable components, APIs, or sophisticated application experiences directly inside SharePoint. Siddharth also argues that generative AI coding tools are changing the traditional assumption that pro-code development necessarily takes longer than low-code development.
THE MODERN SPFx TECHNOLOGY STACK
A modern SPFx project brings together several technologies rather than relying on one framework.Siddharth breaks down the roles of TypeScript, React, Fluent UI, PnPjs, and PnP React Controls. TypeScript provides stronger typing and compile-time checks, while Fluent UI helps custom solutions retain the familiar Microsoft user experience.PnPjs simplifies interaction with SharePoint, Microsoft Graph, and other Microsoft 365 services by replacing repetitive REST request code with reusable abstractions.
STOP REBUILDING CONTROLS THAT ALREADY EXIST
One of the central lessons of the episode is simple: professional development does not mean writing everything yourself.PnP React Controls provide SharePoint-aware and Microsoft 365-aware components for common development requirements. Instead of repeatedly creating the UI, API calls, data binding, and associated logic for components such as file pickers, developers can use established community controls.Siddharth's preferred approach is to check existing capabilities first: use Microsoft functionality when available, then Fluent UI or PnP React Controls where they satisfy the requirement, and create a custom component only when the required functionality does not already exist.
THE FIVE PnP REACT CONTROLS DEVELOPERS SHOULD KNOW
If Siddharth had to choose only five controls, his selection would be People Picker, Taxonomy Picker, List View, File Picker, and Live Persona.These components cover several recurring requirements in enterprise SharePoint applications, including selecting users, working with managed metadata, presenting SharePoint data, selecting or uploading files, and displaying Microsoft 365 user information.
BUILDING REAL APPLICATIONS INSIDE SHAREPOINT
The conversation moves from individual controls to application architecture with the example of a sophisticated project management solution.SharePoint lists can provide the underlying data layer for projects, customers, resources, and tasks, while SPFx can deliver a unified application experience containing dashboards, project views, charts, CRUD operations, role-specific interfaces, task management, and navigation.The result can feel much more like a dedicated business application while remaining embedded inside the SharePoint environment users already know.
SHAREPOINT DOESN'T HAVE TO BE YOUR DATABASE
SPFx applications are not restricted to SharePoint data.Siddharth discusses retrieving information from Dataverse and integrating external systems. When data resides in systems such as Azure SQL, a backend API can provide the secure middle layer between the client-side SPFx application and the database.He also describes a real example where an SPFx web part surfaces Power Automate approvals directly inside SharePoint and allows users to approve or reject requests without moving to another application.
MICROSOFT GRAPH AND SPFx
Microsoft Graph expands SPFx far beyond SharePoint itself.Applications can interact with Microsoft 365 services including OneDrive, Planner, Outlook, meetings, and other resources exposed through Graph. Siddharth explains when SharePoint REST APIs remain appropriate and when Graph becomes the better or necessary option.
SECURITY, PERMISSIONS AND LEAST PRIVILEGE
Security is a major part of professional SPFx development.SPFx solutions calling Microsoft Graph typically operate using delegated permissions and therefore respect the identity and access rights of the currently signed-in user. Requested API permissions also require administrative approval.One of the most common mistakes Siddharth sees is requesting more permissions than the application actually requires. His recommendation is to start with the minimum permissions necessary rather than granting broad access by default.Developers also need to test solutions from the perspective of real users instead of assuming that permissions available during development will also exist in production.
NEVER PUT SECRETS IN CLIENT-SIDE SPFx CODE
Because SPFx executes client-side, sensitive secrets should never be embedded directly into the application code.For scenarios requiring secrets or credentials, Siddharth recommends introducing a backend API that can securely access services such as Azure Key Vault while the SPFx frontend communicates only with that API.
SECURITY REVIEW DOESN'T END WITH YOUR OWN CODE
SPFx relies heavily on the modern JavaScript and npm ecosystem. Organizations therefore need to consider the security and maintenance status of third-party packages as well as their own application logic.Siddharth recommends reviewing dependencies, paying attention to package warnings and vulnerabilities, and including security assessment as part of the deployment process rather than assuming every dependency is safe simply because it is available through npm.
WHY IS YOUR SPFx SOLUTION SO SLOW?
When an SPFx application performs badly, Siddharth starts with the browser's network tools.Developers should examine how many API calls occur during page load, identify unnecessary requests, look for API calls accidentally executed inside loops, and inspect React components for excessive rendering or state changes.A seemingly simple application can generate dozens of requests when data retrieval is implemented inefficiently.
BATCH YOUR REQUESTS
Once unnecessary API traffic has been identified, batching can significantly improve how requests are handled.Instead of sending multiple individual operations from the client, developers can combine appropriate SharePoint operations into batch requests and reduce client-side request overhead.
FROM DEVELOPMENT TO THE SHAREPOINT APP CATALOG
Siddharth also walks through the SPFx deployment process, from packaging the solution into an .sppkg package to deploying it through the SharePoint App Catalog and approving required API permissions.Importantly, deploying a package does not automatically mean installing it everywhere. Organizations can control which SharePoint sites receive the application, and site collection App Catalogs can provide an even narrower deployment scope.
SPFx MEETS COPILOT AND AI AGENTS
SPFx is also moving into the agent era.Siddharth discusses SharePoint Copilot apps and how SPFx can provide interactive user-interface components inside Microsoft 365 Copilot experiences. Instead of returning only text or Markdown, an agent can potentially surface richer interfaces that users can interact with directly.He describes the concept as similar to taking the idea behind Adaptive Cards much further by enabling richer, more customizable application experiences.
THE BIG TAKEAWAY
The strongest SharePoint developers are not necessarily the developers who write the most code.They know when to use SharePoint out of the box, when Power Platform is sufficient, when SPFx provides the necessary flexibility, when Microsoft Graph is required, and when existing Fluent UI and PnP components can eliminate unnecessary development.The goal is not to reinvent another component. It is to combine the Microsoft 365 ecosystem into solutions that are secure, maintainable, performant, accessible, and capable of solving an actual business problem.
ABOUT SIDDHARTH VAGHASIA
Siddharth Vaghasia is a Microsoft MVP, consultant, founder, speaker, blogger, and community contributor specializing in Microsoft 365, SharePoint, Power Platform, Azure, and related technologies. He also discusses his company Binary Roots and its work with customers across multiple international markets. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Transcribed
27 AUG 2026 · For years, SharePoint has carried a reputation for complicated document libraries, outdated intranets, confusing navigation, too many sites, and information that employees simply cannot find. But Microsoft 365 is changing, and Copilot is making SharePoint more important than it has been in a long time.In this episode of the M365 FM Podcast, Mirko Peters talks with Microsoft MVP Marcin Siewnicki about how SharePoint is evolving from a traditional document and intranet platform into a central information and knowledge layer for the modern workplace. Marcin has worked with SharePoint since the SharePoint 2003 era, giving him more than two decades of perspective on how the platform has changed.
WHY SHAREPOINT GOT A BAD REPUTATION
Much of SharePoint’s reputation was created by the way organizations implemented it. Intranets were often designed by IT, heavily customized, difficult to use, and disconnected from what employees actually needed.Instead of continuously gathering feedback and improving the experience, many companies delivered an intranet and expected employees to adapt to it. The result was complex navigation, outdated content, overloaded pages, and systems people avoided whenever possible.
INFORMATION ARCHITECTURE BEFORE TECHNOLOGY
One of the biggest problems is not SharePoint itself but how information is organized inside it. Without a clear information architecture, SharePoint can quickly become a huge shared folder filled with duplicated files, outdated information, unclear ownership, and multiple versions of the same document.Marcin explains why organizations should start with a simple structure, understand their data, remove unnecessary content, and introduce basic metadata. Information architecture does not have to be complicated to make SharePoint significantly easier to use.
BUILDING A MODERN SHAREPOINT WORKPLACE
A modern SharePoint homepage should not become an endless collection of web parts, corporate announcements, applications, and links. It should be simple, personalized, visually clear, and focused on what employees actually need.Communication sites can provide departments and organizations with modern, mobile-friendly publishing experiences without requiring the enormous custom intranet projects that were common in the past. The focus should be on useful information, important applications, relevant documents, straightforward navigation, and strong search capabilities.
SOLVING THE SHAREPOINT NAVIGATION PROBLEM
Navigation remains one of SharePoint’s biggest challenges. Organizations frequently attempt to expose every department, application, resource, and internal page through a single navigation structure.Marcin explains why simpler navigation usually works better. Global navigation should focus on the most important destinations, while detailed navigation can be provided closer to individual departments, sites, and business areas. Otherwise, navigation itself becomes another information problem.
SHAREPOINT AND TEAMS SPRAWL
Microsoft Teams and SharePoint are deeply connected. Every new Team and many Teams channels can introduce additional SharePoint resources, which means uncontrolled Teams creation can quickly become uncontrolled SharePoint growth.Organizations therefore need simple creation processes, templates, basic metadata, lifecycle management, and user education. Employees should understand when they need a Team, when a SharePoint site is sufficient, and what happens behind the scenes when these collaboration environments are created.One useful way of looking at the relationship is to think of Microsoft Teams as the collaboration interface while SharePoint provides much of the document and information layer underneath it.
GOVERNANCE WITHOUT KILLING INNOVATION
Governance does not need to mean preventing employees from using new technology. The goal is to provide enough freedom for people to work effectively while maintaining control over information, permissions, external sharing, security, and lifecycle management.Marcin discusses periodically reviewing inactive SharePoint sites and Teams, identifying environments that are no longer needed, checking ownership, reviewing permissions, and archiving or removing obsolete workspaces.Governance also cannot remain exclusively an IT exercise. Business stakeholders and users need to be involved because overly complicated policies often lead employees to find their own workarounds.
PERMISSIONS, SHARING AND SECURITY
Permissions become increasingly difficult to understand as SharePoint environments grow. Users add colleagues, external partners, vendors, and sharing links over time, but access is rarely reviewed with the same frequency.Regular permission reviews should therefore become part of the governance model. Site and Team owners need responsibility for checking who still requires access and whether external sharing remains necessary.Microsoft Purview sensitivity labels can add another security layer by helping organizations classify and protect information based on its sensitivity.
AUTOMATING SHAREPOINT GOVERNANCE
Governance does not have to be completely manual. PowerShell, Microsoft Graph, Power Automate, Azure Logic Apps, and specialized third-party platforms can automate many administrative and governance processes.Marcin sees PowerShell and Microsoft Graph as fundamental tools for administrators, while Power Automate can provide practical automation for provisioning, approvals, notifications, lifecycle processes, and many other SharePoint scenarios.
MODERN DOCUMENT MANAGEMENT
SharePoint is still heavily associated with documents, and documents remain an important part of the platform. The challenge is managing them properly.Metadata, naming conventions, permissions, versioning, storage management, and sensitivity labels all contribute to a healthier information environment. Versioning deserves particular attention because large documents with many retained versions can consume significant amounts of SharePoint storage.
METADATA VS. FOLDERS
The old SharePoint debate between folders and metadata has not completely disappeared. Marcin argues that metadata remains extremely valuable, but organizations need to keep their metadata models simple.If users need to spend too much time deciding how to classify every document, adoption will suffer. Metadata should support the way employees work rather than creating another administrative task.
COPILOT CAN AUTOMATE METADATA
Copilot changes this equation because AI can reduce the manual work associated with metadata. SharePoint can analyze documents, extract relevant information, and populate columns based on the content.Organizations still need to decide which metadata is useful, but employees no longer necessarily have to enter everything manually. This can make structured information much more practical at scale.
COPILOT CHANGES SHAREPOINT
Copilot significantly lowers the amount of SharePoint knowledge employees need before they can start accomplishing useful work.Users can increasingly create content, work with documents, extract information, generate pages, and interact with organizational knowledge using natural language instead of understanding every technical SharePoint concept.Many of the repetitive tasks traditionally associated with maintaining SharePoint environments can therefore become easier.
FROM FINDING DOCUMENTS TO GETTING ANSWERS
This may be the biggest change of all.Traditional SharePoint experiences required employees to find the right site, navigate to the right library, locate the correct document, open it, and search for the information they needed.Copilot changes that interaction. Instead of only finding documents, employees can increasingly ask questions and receive answers based on information stored across their Microsoft 365 environment.SharePoint therefore starts moving from document storage toward becoming an enterprise knowledge layer.
COPILOT CAN UNDERSTAND MORE THAN OFFICE DOCUMENTS
Organizational knowledge is not limited to Word documents, Excel files, and PowerPoint presentations.Copilot can work with different types of information, including text, HTML, Markdown, SharePoint lists, and content extracted from video transcripts.That means training videos, recorded meetings, presentations, and other media stored within the Microsoft 365 environment can become part of the knowledge available through AI.
WHAT HAPPENED TO SHAREPOINT AGENTS?
SharePoint Agents provide dedicated AI experiences grounded in selected SharePoint sites, document libraries, or information sources.Marcin discusses where these agents fit into the current Copilot landscape and why the concept has become less prominent as Copilot itself gains broader capabilities for interacting directly with SharePoint information.
IS SHAREPOINT REALLY COPILOT’S KNOWLEDGE LAYER?
There is certainly marketing around the idea, but there is also a strong technical reality behind it.Organizations already store enormous amounts of business information in SharePoint and Microsoft Teams. Copilot can use this information, which means the quality of the underlying environment directly affects the quality of the AI experience.Poor permissions, outdated documents, duplicated information, weak metadata, and uncontrolled SharePoint sprawl do not disappear because Copilot has been introduced. They potentially become even more important Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Transcribed
25 AUG 2026 · Microsoft Fabric brings data engineering, analytics, business intelligence, governance and increasingly AI together in one platform. But what does an end-to-end Fabric architecture actually look like when you move beyond individual features and start connecting everything?In this episode of the M365 FM Podcast, Mirko Peters is joined by Amit Chandak [Microsoft Data Platform MVP] for a practical journey through Microsoft Fabric — starting with raw organizational data and ending with trusted information that business users can use to make decisions.
WHY MICROSOFT FABRIC?
Before Fabric, organizations could already build sophisticated analytics architectures using Azure, Power BI and other platforms. The problem wasn't a lack of technology. In many cases, it was the opposite: organizations had too many choices, separate storage technologies, different compute models and multiple copies of essentially the same data.Amit explains how Microsoft Fabric attempts to simplify this architecture by bringing workloads together around shared foundations such as OneLake, common Fabric capacity and the Delta format. Lakehouses, warehouses, Power BI and other Fabric experiences can therefore operate as parts of a broader platform instead of completely isolated services.
ONELAKE AS THE FOUNDATION
OneLake is one of the central concepts behind Fabric. Amit compares it conceptually to OneDrive: instead of every analytics workload creating completely independent storage environments, OneLake provides a virtualized storage foundation across the Fabric tenant.Organizations can still separate data through workspaces, Lakehouses, Warehouses and domains, but those resources exist within a common Fabric storage architecture. This becomes particularly important when organizations want to reduce unnecessary duplication while maintaining security and organizational boundaries.
CENTRALIZED DATA OR DATA MESH?
Fabric doesn't automatically mean putting everything into one giant centralized analytics environment.For smaller organizations, a centralized architecture may still work well. As organizations become larger, Amit sees increasing value in domain-oriented architectures where areas such as sales, finance and purchasing can have their own workspaces and responsibilities.IT can remain responsible for availability, governance and the technical foundation while business domains increasingly take ownership of how their data is analyzed and consumed.
SHORTCUTS INSTEAD OF COPYING DATA
One of the recurring themes throughout the conversation is avoiding unnecessary copies of data.Fabric Shortcuts allow teams to reference data stored elsewhere rather than physically copying it into every environment that needs it. That can apply both inside Fabric and to supported external storage.Amit also explains an interesting architectural benefit of shortcuts: they can help separate workloads across capacities. This can become important when organizations want Power BI consumption workloads isolated from intensive data engineering workloads while still working with the same underlying information.
LAKEHOUSE VS. WAREHOUSE
One of the biggest Fabric architecture questions remains: Should you use a Lakehouse or a Warehouse?A Lakehouse can work with structured and unstructured data and is naturally aligned with Spark. A Fabric Warehouse focuses on structured data and provides the familiar T-SQL experience.Both ultimately use Delta for structured data inside Fabric, which means the decision increasingly comes down to the type of data, preferred technologies and workloads.Organizations with strong SQL teams don't necessarily need to abandon their existing skills. Teams working with very large datasets, advanced engineering scenarios, unstructured information or extensive data science workloads may find the Lakehouse and Spark approach more attractive.
GETTING DATA INTO FABRIC
Once the architecture is defined, organizations still need to bring data into Fabric.Amit walks through several approaches, including Shortcuts, Mirroring, Pipelines, Copy Activity, Copy Jobs, Dataflow Gen2 and notebooks.The right option depends heavily on the source and use case. Dataflow Gen2 remains particularly useful because of its broad connector support and familiar Power Query experience. For Power BI professionals entering Fabric, this can provide a natural starting point before moving toward more engineering-oriented approaches.
WHEN PYSPARK BECOMES IMPORTANT
Power Query and Dataflow Gen2 can work very well for small and medium-sized workloads, but scale changes the equation.For larger transformation workloads, Amit sees significant advantages in Spark-based processing. PySpark and Spark notebooks provide greater flexibility and are designed for distributed processing at scale.SQL and Power BI professionals don't necessarily have to make that transition immediately. Spark SQL can provide a familiar entry point for SQL developers, while many PySpark operations have conceptual similarities to transformations Power Query users already understand.
DO YOU REALLY NEED BRONZE, SILVER AND GOLD?
The Medallion Architecture has become almost synonymous with modern data engineering: Bronze for raw data, Silver for cleaned and transformed data, and Gold for business-ready information.But Amit argues that organizations shouldn't create layers simply because an architecture diagram says they should.If an organization already has excellent master data management and high-quality source data, every intermediate layer may not provide enough value to justify another copy and another transformation step. Where source data quality is inconsistent, however, the traditional Bronze-Silver-Gold structure remains highly valuable.
WHERE SHOULD BUSINESS LOGIC LIVE?
Another important architecture decision is determining where calculations and business rules belong.Amit prefers keeping measures and KPIs in the semantic model where possible because they remain dynamic and easier to change. Heavy row-level calculations across millions of records, however, are generally better handled earlier in the Gold layer.The result is not an either-or decision. A mature Fabric architecture distributes business logic deliberately between the transformation layer and semantic model depending on the type and cost of the calculation.
SEMANTIC MODELS ARE BECOMING MORE IMPORTANT
The semantic model was traditionally viewed primarily as the foundation for Power BI reports. AI is changing that role.Relationships, measures, KPIs and business definitions encoded within a semantic model represent organizational knowledge. That knowledge can increasingly be consumed by experiences beyond traditional dashboards.Amit discusses how Data Agents, ontology-driven solutions and Fabric Apps can build on semantic models. Instead of being merely a Power BI component, the semantic model can become a reusable business layer between organizational data and multiple human or AI-driven experiences.
DOES DAX STILL MATTER IN THE AI ERA?
AI can already generate DAX, SQL, PySpark and other code remarkably quickly. Does that mean professionals no longer need to learn these languages?Amit argues that expertise still matters, particularly when generated solutions don't work correctly or need optimization. AI can dramatically reduce the amount of syntax professionals need to write manually, but understanding the underlying logic remains valuable for debugging and improving what AI creates.Over time, the skill may shift from remembering syntax toward understanding architecture, algorithms, business logic and how to describe requirements precisely.
SECURITY FROM WORKSPACE TO DATA
Putting more organizational data into a unified platform makes security increasingly important.The discussion covers security across multiple layers, beginning with Fabric workspace roles and continuing through item-level access, OneLake security and semantic-model security.For larger organizations, Amit recommends using security groups rather than managing individual users wherever possible. Microsoft Entra provides the identity and group foundation, while Fabric applies those identities and groups to workspaces, items and data access.
FROM DATA ENGINEERING TO BUSINESS DECISIONS
Ultimately, most business users don't care whether their information came through PySpark, a Lakehouse, Delta tables or Dataflow Gen2. They care about getting reliable answers.This is where Amit sees Data Agents, ontology and AI-driven analytics becoming increasingly important.Instead of every user receiving the same predefined dashboard, future analytics experiences could become far more dynamic. Users may interact conversationally with organizational data and eventually have personalized reports or applications generated around their specific questions and responsibilities.
RAPID FIRE
The episode closes with a rapid-fire round.SQL or PySpark? PySpark.Dataflow Gen2 or Notebook? Notebook.DAX or SQL? DAX.Most underrated Fabric feature? Fabric Apps.Biggest Power BI modeling mistake? Many-to-many relationships.And perhaps the most provocative answer of the episode:Are Fabric Apps the new Power BI?Amit's answer: Yes.
THE BIG PICTURE
Microsoft Fabric isn't simply another analytics product to add to the Microsoft stack.Its larger opportunity is connecting ingestion, engineering, storage, Lakehouse and Warehouse architectures, semantic models, security, Power BI and emerging AI experiences into one coherent analytics platform.The technical architecture matters, but the endpoint isn't OneLake, PySpark or even Power BI.The endpoint is trusted information that people — and increasingly AI agents — can turn into better business decisions. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Transcribed
24 AUG 2026 · Enterprise AI is moving beyond chatbots that simply answer questions. The next generation of AI agents can understand business context, connect to enterprise systems, orchestrate workflows, and take action on behalf of users.But building an impressive AI agent demo is easy. Building an agent that works reliably across a global enterprise—with sensitive data, complex business processes, governance requirements, thousands of users, and measurable outcomes—is a very different challenge.In this episode of the M365 FM Podcast, Mirko Peters talks with Manpreet Singh [MVP/MCT] about what it actually takes to build production-ready enterprise AI agents with Microsoft Copilot Studio and the wider Microsoft AI ecosystem.Manpreet explains how organizations are evolving from individual departmental agents toward multi-agent architectures where specialized agents collaborate behind a single interface. The discussion explores when to use Microsoft 365 Copilot, Copilot Studio, and Azure AI Foundry—and how these technologies can work together rather than becoming isolated AI platforms.
FROM ANSWERS TO ACTIONS
The real transformation begins when an AI agent can do more than retrieve information.Using practical examples, Manpreet explains how agents can connect with systems such as Workday, Salesforce, SAP, ServiceNow, Jira, Confluence, and PeopleSoft. Instead of navigating several applications manually, employees can interact with an agent through Microsoft Teams or another conversational interface.An employee requesting leave, for example, could have an agent check available leave, consult HR policies, initiate the request in the underlying HR system, ask the manager for approval, and return the final result—all without the employee opening the individual applications.
KNOWLEDGE, TOOLS AND TRIGGERS
A useful enterprise agent requires more than a good prompt.Manpreet breaks the architecture down into essential elements: knowledge sources that provide organizational context, tools and connectors that allow the agent to interact with enterprise systems, and triggers that determine when processes should begin.SharePoint can become an important knowledge layer, while Power Platform connectors and APIs enable agents to perform actions across business applications.
WHY DATA QUALITY MATTERS
Connecting an agent to twenty years of SharePoint content is not necessarily a good strategy.Manpreet strongly recommends cleaning and curating organizational knowledge before exposing it to AI. Thousands of outdated PDFs, duplicate documents, missing metadata, and obsolete policies can undermine the quality of agent responses.A smaller, carefully maintained knowledge base with current documents, useful metadata, tags, descriptions, and version management can produce significantly better results than simply indexing everything an organization owns.
HUMAN-IN-THE-LOOP AI
Autonomous does not have to mean uncontrolled.For low-risk transactions, organizations may allow an agent to complete an action automatically. Higher-value or sensitive decisions can introduce human approval.Manpreet discusses examples including financial claims, invoice processing, access requests, and infrastructure changes where humans remain part of the decision-making process while AI handles much of the repetitive work surrounding the decision.
GOVERNANCE BEFORE SCALE
As agents gain access to multiple enterprise applications, governance becomes critical.The conversation explores Microsoft Purview, Data Loss Prevention policies, security controls, sensitivity labels, Microsoft Defender, identity, environment strategies, and the importance of establishing an AI Center of Excellence before allowing agent development to expand throughout an organization.Governance should protect enterprise information without creating policies so restrictive that agent performance and usability suffer.
CONTROLLING AGENT SPRAWL
Organizations can quickly move from a handful of experimental agents to hundreds or thousands.Manpreet discusses agent lifecycle management, parent-child agent relationships, centralized inventories, ownership, connected data sources, permissions, consumption, and emerging management capabilities around Agent 365.The goal is to give administrators visibility into which agents exist, who owns them, what they can access, and how they are being used.
THE AI COMMAND CENTER
One of the most interesting concepts discussed in the episode is a unified AI Command Center.Instead of agent creation happening without oversight, organizations can establish a central process for requesting agents, connectors, APIs, MCP integrations, environments, and permissions.The command center can combine approval workflows, auditing, monitoring, governance, usage information, and cost controls—creating a central operational layer for enterprise AI.
OBSERVABILITY AND TROUBLESHOOTING
Traditional applications have logs. Enterprise AI needs observability.When an employee reports that an agent produced an unexpected result, administrators need to understand which knowledge sources were accessed, which tools were called, what actions occurred, and where the process failed.Copilot Studio's tracing and evaluation capabilities can help teams investigate these execution paths and understand how an agent arrived at an outcome.
TESTING NON-DETERMINISTIC AI
Testing AI agents requires a different mindset from testing traditional applications.Manpreet discusses evaluation features, generated test cases, structured datasets, indexing, user feedback, and repeated validation. Organizations should continuously test their agents against realistic questions and expected outcomes rather than deploying an agent once and assuming it will behave correctly forever.
MULTI-AGENT ARCHITECTURES
Instead of forcing employees to find the correct agent for every task, organizations can create an orchestration layer.A primary agent can understand the user's intent and delegate work to specialized agents for HR, travel, marketing, sales, IT, or other functions.The user interacts with one interface while multiple specialized agents operate behind it.
ADOPTION IS WHERE AI ROI HAPPENS
The final challenge is not technical.Organizations can build thousands of sophisticated agents and still fail to generate meaningful business value if employees do not understand when, why, and how to use them.Manpreet argues that adoption, persona-based training, practical use cases, and helping employees integrate agents into their daily work are essential to realizing ROI from Microsoft 365 Copilot and enterprise AI investments.
IN THIS EPISODE
- Microsoft Copilot Studio and enterprise AI agents
- Microsoft 365 Copilot vs. Copilot Studio vs. Azure AI Foundry
- Autonomous agents and action-oriented AI
- Enterprise knowledge grounding
- SharePoint as an AI knowledge source
- Data quality, metadata, and indexing
- Power Platform connectors and enterprise integrations
- Human-in-the-loop architectures
- Multi-agent systems and orchestration
- Microsoft Purview and DLP
- AI security and governance
- Agent lifecycle management
- Agent inventories and Agent 365
- AI Centers of Excellence
- Unified AI Command Centers
- AI observability and tracing
- Agent testing and evaluation
- Hallucination reduction
- AI consumption and cost management
- Enterprise AI adoption and ROI
KEY TAKEAWAY
The next phase of enterprise AI is not simply about adding better chatbots. It is about moving from answers to outcomes.AI agents can connect organizational knowledge, applications, workflows, and business processes—but the fundamentals of enterprise technology remain essential: identity, security, governance, data quality, architecture, testing, observability, cost control, and adoption.And ultimately, the technology only creates value when people actually use it. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Transcribed
23 AUG 2026 · Microsoft 365 Copilot has moved beyond the question of “What can generative AI do?” The harder challenge is now turning AI into something thousands of employees actually use, trust, and derive measurable business value from. In this episode of M365 FM, Mirko Peters talks with Microsoft MVP Christoffer Besler Hansen, Head of AI Workplace at Atea Group, about what it takes to move from a Copilot rollout to a genuine AI-powered workplace. Drawing on experience supporting AI adoption across more than 8,000 employees, Christoffer shares lessons on Microsoft 365 Copilot adoption, training, agents, Copilot Studio, Microsoft Foundry, governance, security, extensibility, FinOps, and measuring business value.
WHAT IS AN AI WORKPLACE?
An AI workplace is much broader than simply giving employees Microsoft 365 Copilot licenses. At Atea, the AI Workplace team is responsible for how more than 8,000 users incorporate AI into their daily work. Microsoft Copilot is a major component, but the strategy also includes Copilot Studio, agentic AI, Foundry, experimentation with other technologies, and—critically—continuous user adoption. The objective is not to deploy one AI product. It is to change how people work with information, applications, and business processes.
TECHNOLOGY MOVES FAST. PEOPLE NEED TIME.
New AI models and capabilities can appear every week. Human working habits don't change at the same speed. Christoffer explains why organizations need to spend substantial effort helping employees feel comfortable changing established workflows. At Atea, this includes training throughout the year, sometimes every other week, with sessions designed for different audiences such as managers, consultants, salespeople, beginners, and advanced users. AI adoption therefore isn't a launch event. It is an ongoing organizational capability.
BUYING 5,000 COPILOT LICENSES ISN'T A STRATEGY
What should happen after an organization purchases thousands of Microsoft 365 Copilot licenses? According to Christoffer, the organization first needs to determine why it purchased them. What is the objective? What should employees accomplish differently? How will the organization support adoption? How will success be measured? Simply assigning licenses and expecting employees to teach themselves isn't enough. Employees already have jobs to perform and cannot realistically follow every weekly change across rapidly evolving AI products.
WHY EARLY COPILOT ADOPTION OFTEN DROPS
AI naturally generates curiosity. When users initially received Copilot without structured adoption support, Christoffer observed strong engagement for approximately the first four weeks. Employees experimented with the technology. But when they struggled to turn those experiments into new working habits, usage declined. After structured training was introduced, users were more likely to continue using Copilot over time—and began asking for additional training as the products evolved. Initial excitement gets people through the door. Continuous education helps keep them there.
HOW DO YOU MEASURE COPILOT ROI?
One of the hardest enterprise AI questions is determining whether Copilot is actually creating value. Usage alone isn't enough. An employee opening Copilot 50 times doesn't necessarily mean the organization has become more productive. Christoffer argues that organizations need to identify what matters to their particular business and then measure whether AI improves those outcomes. That could include completing work faster, handling more customer cases, improving quality, or increasing business capacity.
MEASURE OUTPUT, NOT JUST AI USAGE
One example discussed in the episode involves an employee who previously handled two cases simultaneously but could use AI to work across six while still receiving better customer feedback. That represents something more meaningful than a Copilot usage statistic. The employee is producing more output while maintaining or improving quality. The right KPI therefore depends on what the organization actually produces. AI metrics should ultimately connect with business metrics.
COPILOT AS A THINKING PARTNER
Meetings were one of the earliest areas where Microsoft 365 Copilot delivered obvious value. Transcription, summaries, and meeting intelligence can reduce administrative effort. But Christoffer highlights another important pattern: using AI as a thinking partner. Instead of asking AI to do all the thinking, start with your own ideas. Speak or dictate those thoughts. Let Copilot structure them. Review the result. Give feedback. Iterate until you have something useful. This approach can save time while simultaneously improving the quality of emails, presentations, documents, and other knowledge work.
RESEARCH AGENT CHANGES KNOWLEDGE WORK
Christoffer highlights Microsoft's Research Agent as one of the particularly valuable additions to Copilot. For large projects involving significant amounts of information, an agent capable of working through numerous sources can dramatically reduce research effort. It can also help users find information they may previously have struggled to discover manually. Importantly, users can review the underlying sources and verify whether the resulting information is correct.
YOUR AI IS ONLY AS GOOD AS YOUR INFORMATION
Enterprise AI quickly exposes existing information-management problems. Organizations need to think about how information is structured across SharePoint, OneDrive, CRM systems, and other repositories. Microsoft Purview can play an important role in identifying and protecting confidential information. Retention policies also matter because outdated documents can lead AI toward outdated answers. Simply giving an agent access to more information doesn't automatically make it better. Sometimes the correct approach is to clean the data before connecting the agent.
WHEN SHOULD YOU BUILD AN AGENT? ㅤ Christoffer recommends encouraging employees to start thinking about potential agent use cases early. Initially, organizations may create many simple agents that primarily retrieve information. Some will provide little long-term value. But experimentation changes how employees think about automation. The next maturity step is building agents that don't merely answer questions but perform actions—sending messages, updating CRM systems, triggering processes, or interacting with other applications. Eventually, organizations can move toward more autonomous agents working alongside employees.
AGENT BUILDER VS COPILOT STUDIO
Not every employee needs to begin with Copilot Studio. Christoffer sees many non-technical employees using Agent Builder directly inside Copilot to create simpler agents. More technical users move toward Copilot Studio when they require additional capabilities. This can create a useful progression: Idea → Simple Agent → Validation → Copilot Studio → Advanced Enterprise Agent An employee can prove the concept without becoming a professional developer, then involve technical specialists when the solution needs to become more sophisticated.
FROM ANSWERS TO ACTIONS
An HR agent answering “How many vacation days do I have?” is useful. An agent that can actually book next Friday as vacation represents a fundamentally different capability. This transition from information retrieval toward actions changes the architecture and security requirements surrounding AI. Christoffer expects users to interact less directly with traditional application interfaces as agents increasingly perform tasks on their behalf. Instead of navigating several administrative screens, users may simply describe the desired outcome to an agent.
AGENT SECURITY BECOMES CRITICAL
Once agents can take actions, organizations need strong guardrails. What can the agent do automatically? What requires explicit approval? What can it delete? Which systems can it access? What permissions does it receive? Christoffer emphasizes least privilege and approval controls, particularly when agents interact with administrative environments. Giving an autonomous agent Global Administrator privileges and allowing it to operate without restrictions would create obvious risks. The more capable agents become, the more important their permission architecture becomes.
ENTERPRISE AGENTS NEED AN INTAKE PROCESS
A personal agent used by one employee is different from an agent deployed to thousands of users. Enterprise-wide agents need quality control. Organizations should review instructions, permissions, connectors, integrations, and data access before allowing large numbers of employees to use them. Christoffer suggests establishing an intake process where employee ideas can be evaluated. Some agents may be returned to their creators for improvement. Strategically important ideas can instead be developed by a dedicated internal agent team.
COPILOT EXTENSIBILITY AND AGENT 365
The conversation also explores Copilot extensibility and the Agent 365 SDK. Christoffer describes experimenting with personal agents that have their own identities in Microsoft Entra. Such an agent could appear within an organizational structure, have its own email and Teams presence, and receive carefully controlled permissions. His example involves building an agent that can act as a personal assistant and potentially answer appropriate questions when he is away from work. The important architectural shift is that the agent begins looking less like a chatbot and more like another identity participating in the organization. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
Supporters Club
5 NOV 2025 · Are your Microsoft Copilot logs silently exposing sensitive data? In this episode of M365.FM, we reveal how unmanaged Copilot activity logs can create serious compliance risks, storage bloat, and security blind spots in your Microsoft 365 environment.
What you will learn in this episode:
- Why unmanaged Microsoft Copilot logs are a hidden security and compliance risk
- - How to use Microsoft Purview to capture, audit, and control Copilot activity logs
- - Step-by-step: enabling auditing and setting precise scopes for Copilot log management
- - How to establish data retention strategies and automate log exports
- - How to limit access to sensitive Copilot logs and avoid common configuration mistakes
- - How to operationalize workflows and build repeatable discovery processes under pressure
- - Practical tips to take control of Copilot logs for accountability, peace of mind, and data protection
Whether you are a Microsoft 365 admin, security professional, or compliance officer, this episode gives you actionable strategies to stop data leakage through Copilot logs before it becomes a costly problem.
About the Host:
Mirko Peters is the host of M365.FM, a podcast dedicated to modern work, security, and productivity with Microsoft 365. With deep expertise in Microsoft cloud technologies, AI, and enterprise security, Mirko shares practical insights to help IT professionals and business leaders get the most out of their Microsoft 365 investment.
Subscribe to M365.FM for weekly episodes on Microsoft 365, Copilot, Power Platform, Microsoft Purview, and enterprise security best practices.
#MicrosoftCopilot #Microsoft365Security #CopilotLogs #MicrosoftPurview #DataProtection #ComplianceManagement #M365Admin #EnterpriseAI #CloudSecurity #M365FM Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-a-microsoft-mvp-podcast-by-mirko-peters--6704921/support.
M365.FM is a podcast about Microsoft 365, Microsoft Copilot, AI, Modern Work, security, governance, Power Platform, Azure, and the technologies shaping the future of work.Hosted by Microsoft MVP Mirko Peters,...
show more
M365.FM is a podcast about Microsoft 365, Microsoft Copilot, AI, Modern Work, security, governance, Power Platform, Azure, and the technologies shaping the future of work.Hosted by Microsoft MVP Mirko Peters, M365.FM brings together Microsoft MVPs, Microsoft employees, product experts, architects, developers, and community leaders from around the world.Each episode goes beyond announcements and hype to explore what Microsoft technologies mean in practice. From Microsoft 365 Copilot and AI agents to Teams, SharePoint, Power Platform, Microsoft Fabric, Entra, Purview, security, governance, adoption, and automation, M365.FM focuses on real-world experience, implementation, strategy, and lessons learned.Expect expert interviews, technical deep dives, practical explainers, and conversations with people building, implementing, and shaping the Microsoft ecosystem.If you work with Microsoft 365, Copilot, AI, Modern Work, or the Microsoft Cloud, M365.FM helps you understand what matters, what works, and what is coming next.Hosted by Mirko Peters, Microsoft MVP.
show less
Information
| Author | Mirko Peters (M365 Consultant) |
| Organization | m365 FM |
| Categories | Tech News , Technology , How To |
| Website | www.m365.fm |
| mirko.peters@m365.show |
Copyright 2026 - Spreaker Inc. an iHeartMedia Company
