M365.FM - Modern work, security, and productivity with Microsoft 365

Mirko Peters - Founder of m365.fm, m365.show and m365con.net

Welcome to the M365.FM — your essential podcast for everything Microsoft 365, Azure, and beyond. Join us as we explore the latest developments across Power BI, Power Platform, Microsoft Teams, Viva, Fabric, Purview, Security, and the entire Microsoft ecosystem. Each episode delivers expert insights, real-world use cases, best practices, and interviews with industry leaders to help you stay ahead in the fast-moving world of cloud, collaboration, and data innovation. Whether you're an IT professional, business leader, developer, or data enthusiast, the M365.FM brings the knowledge, trends, and strategies you need to thrive in the modern digital workplace. Tune in, level up, and make the most of everything Microsoft has to offer. M365.FM is part of the M365-Show Network. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.

  1. 6시간 전

    The Death of the Chatbot: Why Your Dataverse Strategy Is Broken

    Microsoft Copilot has transformed how organizations interact with AI, making conversational experiences more accessible than ever. But while chat-based AI delivers immediate productivity gains, it does not provide the architectural foundation required for enterprise-scale autonomous agents. As organizations deploy more AI solutions across departments, they quickly encounter governance challenges, identity issues, fragmented integrations, and uncontrolled costs. Copilot is an excellent interface—but it is only one layer of a much larger AI ecosystem. AGENT IDENTITY, GOVERNANCE, AND SECURITY FOR ENTERPRISE AI One of the biggest challenges in enterprise AI is identity. Many organizations still allow AI agents to operate under shared service accounts or even employee credentials, making auditing nearly impossible. Every autonomous agent should have its own dedicated identity, least-privilege permissions, and complete traceability. Combined with centralized governance, organizations gain full visibility into who—or what—accessed sensitive data, ensuring compliance with standards such as GDPR, SOC 2, and industry-specific regulations.  FROM RAG TO ONTOLOGIES: BUILDING AGENTS THAT UNDERSTAND BUSINESS CONTEXT Traditional Retrieval-Augmented Generation (RAG) systems retrieve documents and generate answers based on matching text. While useful, they rarely understand how a business actually operates. Agent Mesh architectures replace document-centric reasoning with ontologies that model customers, products, suppliers, policies, and business relationships. Instead of searching for words, AI agents reason over structured knowledge, dramatically improving accuracy, consistency, and decision-making across the enterprise.  THE AI LANDING ZONE: CENTRALIZED CONTROL FOR AGENT MESH Scaling dozens or even hundreds of AI agents requires more than good prompts. Organizations need a dedicated AI Landing Zone that combines identity management, governance policies, model gateways, observability, cost controls, and centralized approval processes. Every model request flows through a governance layer where security, regional compliance, budget limits, and policy enforcement are applied automatically. This approach transforms isolated AI projects into a standardized enterprise platform capable of supporting large-scale autonomous operations.  AGENT 365, OBSERVABILITY, AND FINOPS FOR RESPONSIBLE AI Managing AI at scale requires complete operational visibility. A centralized control plane such as Agent 365 enables organizations to inventory agents, monitor usage, assign ownership, retire unused "ghost agents," and analyze every model invocation. Combined with comprehensive observability and FinOps practices, businesses can optimize token consumption, enforce budgets, detect abnormal behavior, and maintain continuous compliance while significantly reducing operational costs.  THE FUTURE OF MICROSOFT AI: FROM COPILOT TO THE AGENT MESH The next generation of enterprise AI is no longer about individual chatbots—it is about interconnected, governed, autonomous systems working together. Organizations that invest early in Agent Mesh architectures, centralized governance, ontology-driven reasoning, secure identities, and AI operating platforms will be able to scale hundreds of intelligent agents safely and efficiently. The future belongs to businesses that treat AI not as a feature, but as enterprise infrastructure capable of supporting continuous automation, intelligent decision-making, and long-term digital transformation. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.

    The Death of the Chatbot: Why Your Dataverse Strategy Is Broken
  2. 11시간 전

    The Future of IT Is Agentic: Inside Windows 365, Intune & Microsoft's AI Vision with Christiaan Brinkhoff

    Christiaan Brinkhoff shares the remarkable career path that took him from speaking at community events and writing technical blogs to becoming one of the key people behind Microsoft's modern cloud desktop strategy. After joining FSLogix, which was later acquired by Microsoft, he helped shape Azure Virtual Desktop before becoming part of the secret development team behind Windows 365. He discusses working in Redmond during one of the most transformative periods in Microsoft's history, contributing to innovations including Windows 365 Boot, Windows 365 Switch, the Windows App, multiple patents, and the evolution of Cloud PCs. Today, he continues driving innovation as VP of Product at Nerdio, helping organizations simplify enterprise endpoint management. WHY CLOUD PCS ARE BECOMING THE FUTURE OF ENTERPRISE COMPUTING Cloud PCs are no longer just a niche virtualization technology. Christiaan explains how Windows 365 fundamentally changes enterprise computing by moving Windows into the cloud while maintaining the familiar user experience. Instead of thinking about remote desktops as a complex virtualization platform, organizations can now manage Cloud PCs through Microsoft Intune just like traditional physical devices. This dramatically lowers the barrier to adoption while making remote work, device replacement, and endpoint security significantly easier to manage.  AZURE VIRTUAL DESKTOP VS. WINDOWS 365 One of the biggest discussions in modern endpoint management is understanding where Azure Virtual Desktop ends and Windows 365 begins. Christiaan explains that Azure Virtual Desktop remains the highly customizable Platform-as-a-Service offering for organizations needing maximum flexibility, while Windows 365 delivers a fully managed Software-as-a-Service experience where Microsoft handles much of the underlying complexity. Rather than replacing each other, both services complement one another, allowing organizations to choose the right solution based on workloads, management capabilities, and business requirements.  HOW COVID ACCELERATED THE CLOUD PC REVOLUTION The pandemic completely transformed the adoption of virtual desktops. Christiaan reflects on how Azure Virtual Desktop evolved from a relatively small service into one of Microsoft's fastest-growing enterprise platforms almost overnight. Organizations suddenly needed secure remote access for thousands of employees, and Microsoft's virtualization technologies became a critical foundation for enabling remote work around the world. This massive adoption also created the demand for an even simpler cloud-native experience, ultimately accelerating the development and success of Windows 365.  THE STORY BEHIND WINDOWS 365 BOOT AND WINDOWS 365 SWITCH Few people know the design decisions behind some of Windows 365's most innovative features. Christiaan explains how Windows 365 Boot was created to remove the complexity of traditional virtual desktop logins by allowing users to boot directly into a Cloud PC. He also shares how Windows 365 Switch enables users to seamlessly move between their local device and their Cloud PC as naturally as switching between Windows virtual desktops. Both features were designed to hide technical complexity and create an experience that feels completely native to Windows users.  WHY MICROSOFT INTUNE BECAME THE FOUNDATION OF WINDOWS 365 One of the smartest strategic decisions Microsoft made was integrating Windows 365 directly into Microsoft Intune rather than building an entirely separate management platform. Christiaan explains how this allows IT administrators to manage Cloud PCs using the same policies, security settings, compliance controls, and deployment processes they already use for physical devices. This unified management experience significantly reduced adoption barriers and made Windows 365 attractive to organizations of every size. FROM COPILOT TO AGENTIC AI Artificial intelligence is rapidly evolving beyond simple chat assistants. Christiaan believes today's Copilot experiences represent only the first generation of enterprise AI. The next stage is Agentic AI, where multiple specialized AI agents collaborate, automate workflows, communicate with systems, and execute complex business processes with minimal human intervention. Instead of interacting with one assistant, organizations will increasingly orchestrate entire teams of AI agents working together behind the scenes. WHY THE FUTURE OF AI HAS ALMOST NO USER INTERFACE One of the most thought-provoking ideas from the conversation is Christiaan's belief that the best AI interface may ultimately become almost invisible. Rather than navigating traditional menus and dashboards, users will simply describe what they want to accomplish while AI dynamically generates reports, dashboards, code, documentation, or management actions in real time. The focus shifts away from software interfaces toward outcomes, making enterprise software significantly more intuitive. REINVENTING ENDPOINT MANAGEMENT WITH AI Endpoint management is entering an entirely new era. Christiaan describes how AI will increasingly automate repetitive administrative work such as policy deployment, reporting, troubleshooting, compliance monitoring, and device lifecycle management. Instead of manually clicking through management consoles, IT administrators will interact with intelligent systems capable of understanding natural language and performing complex operations securely on their behalf.  WHAT NERDIO IS BUILDING FOR THE AI ERA At Nerdio, Christiaan is helping develop the next generation of endpoint management powered by AI. Rather than replacing Microsoft Intune, Nerdio focuses on simplifying complex management tasks, integrating AI-driven automation, and enabling administrators to manage Windows 365, Azure Virtual Desktop, and traditional endpoints through intelligent workflows. The long-term vision is an adaptive management platform capable of understanding business intent rather than simply executing predefined configuration steps.  THE RISE OF THE AGENTIC USER One of the most fascinating concepts discussed is Microsoft's emerging idea of the "Agentic User." Instead of every action being performed directly by a human administrator, organizations will increasingly authorize AI agents to execute tasks on behalf of users while respecting permissions, RBAC controls, and governance policies. This creates digital workers capable of securely performing routine administrative work while humans remain responsible for oversight and strategic decisions.  AI TROUBLESHOOTING IS CLOSER THAN YOU THINK When discussing the future of technical support, Christiaan explains that AI already has the capability to analyze logs, identify patterns, detect root causes, and recommend solutions much faster than human administrators. The remaining challenge is securely connecting these AI capabilities with enterprise management systems so that intelligent troubleshooting can move beyond recommendations and begin performing autonomous remediation where appropriate. WHY  THE NEXT OPERATING SYSTEM MAY LOOK COMPLETELY DIFFERENT According to Christiaan, Windows itself is beginning a significant transformation. As increasingly powerful NPUs and AI accelerators become standard hardware, more AI workloads will execute locally instead of exclusively in the cloud. Combined with advances in AI-first interfaces, the traditional desktop may gradually evolve into a much more conversational, adaptive, and context-aware operating environment where users interact primarily through intelligent assistants instead of conventional applications.  GRAPH API, MCP, AND THE FUTURE OF ENTERPRISE AUTOMATION Graph API remains a critical foundation for Microsoft's ecosystem, but Christiaan expects Model Context Protocol (MCP) and AI orchestration technologies to become increasingly important. Rather than replacing Graph, MCP can build upon it, allowing intelligent agents to securely interact with enterprise systems through standardized interfaces while dramatically expanding automation possibilities across Microsoft 365 and beyond.  WHY COMMUNITY STILL MATTERS Despite working on some of Microsoft's most important cloud technologies, Christiaan repeatedly returns to one message: everything started with community. Blogging, presenting at user groups, sharing knowledge, and helping others opened career opportunities that ultimately led to Microsoft. He encourages listeners to contribute to the community, continuously learn, and remain curious because those investments often create opportunities that cannot be planned in advance RAPID FIRE INSIGHTS During the quick-fire round, Christiaan shares several personal preferences and predictions: Windows 365 over Azure Virtual DesktopMicrosoft Intune over Configuration ManagerAI Agents over traditional Copilot assistantsGraph API over PowerShellMicrosoft Teams over OutlookDiscipline as the most important success habit Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.

    The Future of IT Is Agentic: Inside Windows 365, Intune & Microsoft's AI Vision with Christiaan Brinkhoff
  3. 14시간 전

    How Do You Successfully Deploy Microsoft 365 Copilot Across an Enterprise?

    A successful Microsoft 365 Copilot deployment begins long before licenses are assigned. Organizations should first define measurable business outcomes, identify high-value use cases, and secure executive sponsorship across IT, security, finance, and business leadership. Rather than treating Copilot as another software rollout, enterprises need an AI operating model that aligns governance, adoption, security, and ROI with real business processes. Starting with targeted scenarios creates faster wins while reducing risk and establishing a repeatable framework for future AI initiatives. ASSESS MICROSOFT 365 READINESS BEFORE ENABLING COPILOT One of the biggest mistakes organizations make is assuming their Microsoft 365 tenant is AI-ready simply because they own the licenses. Before deployment, businesses should assess identity management, SharePoint permissions, Teams collaboration spaces, OneDrive sharing, Microsoft Entra ID, Conditional Access, and overall information architecture. Existing oversharing, outdated permissions, unmanaged guest accounts, and poor data ownership become significantly more visible once Copilot can surface enterprise knowledge through natural language. A structured readiness assessment identifies critical risks before they become security incidents during rollout.  SECURE YOUR DATA WITH GOVERNANCE, PERMISSIONS, AND MICROSOFT PURVIEW Microsoft 365 Copilot never creates new permissions—it simply works with the permissions users already have. That makes governance, Microsoft Purview, sensitivity labels, Data Loss Prevention, lifecycle management, and permission cleanup essential parts of every deployment. Organizations should prioritize high-risk content, establish clear ownership of SharePoint sites and Teams, implement strong information protection policies, and continuously review access rights. AI success depends as much on data quality and governance as it does on the underlying technology.  RUN A CONTROLLED COPILOT PILOT BEFORE SCALING ACROSS THE ENTERPRISE Enterprise AI should expand through carefully planned pilot programs instead of company-wide deployments. Successful pilots focus on repeatable business workflows, measurable productivity improvements, and clearly defined success criteria. Business owners, IT, security, and finance should jointly evaluate business outcomes, adoption rates, governance findings, and user feedback before approving additional rollout phases. Every pilot should generate practical lessons that improve future deployments rather than simply proving that Copilot can generate content.  DRIVE MICROSOFT 365 COPILOT ADOPTION WITH CHANGE MANAGEMENT Technology alone does not transform an organization—people do. Successful Copilot adoption requires executive communication, role-based enablement, workflow-specific training, AI champions, ongoing coaching, and continuous learning. Employees need practical guidance on when to trust Copilot, when human review remains mandatory, and how AI supports rather than replaces professional judgment. Measuring adoption should focus on changed business behavior and improved workflows instead of simple prompt counts or login statistics.  MEASURE COPILOT ROI AND BUILD A SCALABLE ENTERPRISE AI PLATFORM The true return on Microsoft 365 Copilot comes from measurable business improvements rather than AI usage alone. Organizations should track workflow efficiency, quality improvements, reduced rework, employee productivity, governance maturity, and financial outcomes across every deployment phase. A successful rollout creates more than a productive workforce—it establishes the governance, architecture, operating model, and organizational experience required to scale future AI capabilities such as Copilot Studio, AI agents, Microsoft Graph integrations, and enterprise automation. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.

    How Do You Successfully Deploy Microsoft 365 Copilot Across an Enterprise?
  4. 1일 전

    From Pilot to Production: Building Enterprise AI That Actually Delivers with Leon Gordon [MVP]

    Leon Gordon explains why most enterprise AI initiatives never reach production and introduces the concept of the Pilot Tax—the hidden cost organizations pay when AI projects remain stuck in proof-of-concept mode. He shares practical strategies for moving from experimentation to measurable business outcomes through governance, Microsoft Fabric, and structured AI adoption. FROM FOOTBALL TO MICROSOFT MVP Leon shares his unconventional career journey, from leaving school early to pursue professional football to becoming a five-time Microsoft MVP, founder of Onyx Data, and one of the leading voices in Microsoft Fabric and enterprise AI. His story demonstrates how continuous learning and real-world experience can outperform traditional career paths.  BUILDING AI THAT DELIVERS BUSINESS VALUE Rather than focusing on flashy AI demonstrations, Leon explains why organizations must begin with measurable business outcomes. Every AI initiative should start by defining success metrics, expected ROI, and governance requirements before writing a single prompt or deploying an agent. WHY  MOST AI PROJECTS FAIL Despite billions being invested worldwide, most generative AI projects never reach production. Leon explores the biggest reasons behind these failures, including weak governance, poor data quality, unrealistic expectations, insufficient testing, and a lack of long-term strategy. He argues that organizations often rush to implement AI before preparing the necessary foundations. THE PILOT TAX EXPLAINED Leon introduces his Pilot Tax methodology, designed to help organizations escape endless proof-of-concept cycles. By focusing on small, measurable Proof of Value projects instead of isolated pilots, companies can validate business impact quickly and create a structured path toward production-ready AI.  MICROSOFT FABRIC AS THE AI FOUNDATION Microsoft Fabric is more than a data platform. Leon explains how it unifies data engineering, analytics, semantic models, AI, real-time intelligence, and application development into a single ecosystem. This dramatically simplifies enterprise architecture while accelerating AI adoption across organizations. FABRIC APPS AND THE FUTURE OF BUSINESS APPLICATIONS Fabric Apps represent one of Microsoft's newest innovations. Leon discusses how they bring application development directly into the Fabric ecosystem, enabling developers to build AI-powered business applications that interact seamlessly with semantic models, analytics, and enterprise data.  GOVERNANCE IS THE REAL COMPETITIVE ADVANTAGE Strong governance is the difference between successful AI deployments and expensive failures. Leon explains why governance must cover security, permissions, ownership, data quality, lineage, metadata, compliance, and continuous monitoring from day one instead of being added later.  WHY METADATA AND MICROSOFT PURVIEW MATTER Metadata often receives little attention until organizations begin implementing AI. Leon explains how Microsoft Purview helps organizations catalog, classify, govern, and secure enterprise data while making it easier for AI systems to understand business context and maintain trust in generated answers.  THE GROWING IMPORTANCE OF SEMANTIC MODELS Semantic models are becoming one of the most valuable assets in modern data platforms. Leon explains how they provide business context, reusable calculations, relationships, and definitions that enable AI agents to deliver accurate, explainable, and trustworthy business insights.  GOVERNANCE SHOULD NEVER WAIT Many organizations prioritize dashboards before governance, promising to "fix it later." Leon argues this almost always creates technical debt. Instead, governance should be embedded throughout the development lifecycle so organizations can deliver value quickly without sacrificing security or maintainability.  INTRODUCING FABOPS Leon presents FabOps, his governance platform for Microsoft Fabric. It provides centralized monitoring, governance, cost management, observability, best-practice validation, performance insights, executive reporting, and FinOps capabilities across an organization's entire Fabric estate.  AI, COPILOT, AND FOUNDRY Copilot is an excellent productivity assistant, but Leon believes organizations unlock far greater value through Azure AI Foundry and intelligent multi-agent architectures. As AI matures, businesses will increasingly orchestrate multiple specialized agents rather than relying on a single assistant experience.  THE FUTURE OF DATA PROFESSIONALS AI will not replace data engineers or analysts—it will amplify them. Leon explains how autonomous engineering agents can dramatically accelerate development while human experts continue to provide architecture, governance, validation, and strategic decision-making. Future professionals will supervise AI rather than compete with it.  QUICK FIRE INSIGHTS During the rapid-fire round, Leon shares his personal favorites: Power BI over Fabric Apps (for now)Coffee over tea or energy drinksCopilot over traditional BI workflowsData Lake over Data WarehouseMicrosoft Fabric as his favorite Microsoft productProfit First as a must-read business bookCommunity as one of the most valuable assets in techContinuous learning as the most important skill for every IT professionalTea and crumpets as the classic British choiceWHAT'S NEXT FOR ONYX DATA Leon closes by sharing his vision for Onyx Data: helping organizations build governed, production-ready AI solutions that generate measurable business value using Microsoft technologies. As enterprise AI continues to evolve, his mission remains focused on turning innovation into real-world outcomes. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.

    From Pilot to Production: Building Enterprise AI That Actually Delivers with Leon Gordon [MVP]
  5. 1일 전

    Microsoft Purview is a Trap: The Hard Truth About Data Governance

    Microsoft Purview is included with many Microsoft 365 subscriptions, making it incredibly easy to enable. That convenience is also its biggest danger. Because there is no procurement process or large implementation project, many organizations activate Purview without defining clear business goals, ownership, or governance. The result is often a catalog filled with thousands of scanned assets, confusing permissions, and business users who abandon the platform after their first experience. This episode explains why Purview itself is not the problem—the real challenge is how organizations approach data governance. Governance must begin with business objectives, ownership, and change management before any scans are executed or collections are created. STOP BUILDING A CATALOG — START SOLVING BUSINESS PROBLEMS One of the biggest mistakes organizations make is attempting to catalog their entire data estate from day one. Instead of asking, "What data do we have?", they should ask, "What business question are we trying to answer?" Every successful Microsoft Purview deployment should begin with a single, measurable use case such as fraud detection, customer churn prediction, or regulatory reporting. That single question determines which data sources need to be scanned, who should own the data, which governance domain is required, and what success looks like. Building one valuable data product first creates trust, enables rapid feedback, and provides a repeatable blueprint for future governance initiatives. A focused rollout consistently delivers better adoption than a large-scale "Big Bang" implementation. DESIGNING MICROSOFT PURVIEW FOR SCALE The episode provides a deep architectural walkthrough of Microsoft Purview's governance model, explaining the four permission layers that control access: the Tenant Layer, the Data Map, the Unified Catalog, and Governance Domains. Rather than assigning permissions directly to individuals, organizations should package permissions into role-based Microsoft Entra groups aligned with real business personas. The discussion also covers how to organize collections around business domains instead of technical platforms, why governance domains should mirror business ownership, and how data products become the bridge between raw technical assets and meaningful business outcomes. By structuring Purview around people, business processes, and ownership rather than databases and technologies, organizations create a catalog that employees can actually understand and use. DATA PRODUCTS, OWNERSHIP, AND THE MEDALLION ACCOUNTABILITY MODEL Governance only succeeds when ownership is clearly defined. The episode explains how data products bring together assets from multiple platforms under a single business purpose, complete with owners, glossary terms, policies, and approval workflows. It also explores how accountability shifts throughout a modern data platform using the Medallion Architecture. Bronze data remains the responsibility of source system owners, Silver data belongs to engineering teams responsible for transformations, and Gold data becomes the responsibility of business-facing data product owners. Explicit ownership at every stage eliminates ambiguity during audits, improves trust in analytics, and ensures someone is always accountable when business-critical data or AI models produce unexpected results. SECURING MICROSOFT PURVIEW WITHOUT CREATING CHAOS Because Microsoft Purview administrators can elevate their own permissions and control nearly every aspect of the platform, privileged access requires special attention. The episode explains why Privileged Identity Management (PIM) should always protect high-privilege roles using just-in-time access, approval workflows, multi-factor authentication, limited activation windows, and full auditing. Beyond security, the rollout strategy itself determines long-term success. Organizations should begin with one governance domain, one business question, one data product, and one pilot audience before expanding. The episode concludes by highlighting the most common implementation failures—including permission sprawl, missing ownership, inconsistent reader permissions, excessive scanning, and poor collection design—and provides practical recommendations for avoiding each of them while building a scalable, business-driven Microsoft Purview governance strategy. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.

    Microsoft Purview is a Trap: The Hard Truth About Data Governance
  6. 2일 전

    From Excel Expert to Microsoft MVP: Empowering Millions with Data, Dashboards & AI with Karen Abecia [Microsoft MVP]

    aren Abecia shares the remarkable journey that transformed a passion for Microsoft Excel into a global career as one of the world's best-known Excel educators. She explains how discovering creative spreadsheet design early in her career led her to help thousands of professionals improve their work, build confidence, and communicate data more effectively. Her story demonstrates that technical expertise combined with genuine passion can create opportunities far beyond traditional career paths. WHY EXCEL CHANGED HER LIFE For Karen, Excel represents much more than software. It gave her financial independence, allowed her to support her family, opened international opportunities, and became a tool for empowering others. She even has two Excel tattoos to symbolize how profoundly the application influenced both her personal and professional life. Rather than viewing Excel as spreadsheets, she sees it as a platform that gives people confidence, recognition, and career growth.  LEARNING WITHOUT A TRADITIONAL EDUCATION Karen discusses building her career without a university degree and explains why continuous learning has always been essential. She believes that formal education is only one path to success and encourages people to study what genuinely excites them. Her philosophy is simple: lifelong curiosity matters more than traditional credentials.  IS EXCEL REALLY DYING? Despite years of headlines claiming that Excel is becoming obsolete, Karen strongly disagrees. She argues that most people predicting Excel's demise do not truly understand how widely it is used across businesses worldwide. Instead of worrying about these predictions, she focuses on helping people solve real problems with the tools they already rely on every day.  THE SECRET OF A GREAT DASHBOARD Creating dashboards is no longer just about technical skills. Karen believes AI can generate standard dashboards for almost anyone, making creativity and presentation more valuable than ever. She explains that outstanding dashboards create a genuine "wow effect" through thoughtful design, visual storytelling, layout, colors, alignment, and attention to detail rather than simply displaying charts and numbers.  COMMON DASHBOARD MISTAKES Many users focus entirely on calculations while overlooking presentation. Karen explains that inconsistent alignment, poor spacing, mismatched colors, incorrect font sizes, and even spelling mistakes can significantly reduce a dashboard's impact. Small visual improvements often make a much larger difference than adding additional formulas or charts.  TEACHING MILLIONS THROUGH SIMPLICITY Having trained more than 15,000 students, Karen believes effective teaching is not about demonstrating advanced technical knowledge. Instead, it is about helping people become more productive and confident using practical techniques they can immediately apply. Her students come from virtually every industry because almost anyone using a computer can benefit from Excel.  EXCEL, AI, AND THE FUTURE OF PRODUCTIVITY Karen discusses how AI is changing Excel workflows and why she actively experiments with multiple AI assistants, including Microsoft Copilot and Claude. Rather than expecting AI to replace expertise, she uses it to accelerate her own creative process while maintaining her personal design style. She also shares her growing interest in Copilot Agents and Microsoft's rapidly evolving AI ecosystem.  EXCEL VS. POWER BI Rather than viewing Excel and Power BI as competitors, Karen now considers them complementary tools. Different organizations require different solutions depending on their size, budget, and reporting needs. While Excel remains her preferred environment for flexibility and creativity, she recognizes that Power BI provides capabilities Excel cannot easily replace.  BUILDING A COMMUNITY THROUGH EMPATHY Karen believes great teachers never make students feel unintelligent. She reflects on early experiences where technical experts made her feel inadequate and explains how those moments shaped her teaching philosophy. Every class is built around patience, encouragement, and making learners feel capable regardless of their experience level.  AUTHENTICITY OVER PERFECTION One of Karen's biggest lessons is that people connect with authenticity rather than perfection. She openly embraces mistakes, especially when speaking English, and believes showing vulnerability creates stronger relationships with students. Instead of trying to appear flawless, she encourages others to simply be themselves.  BECOMING A MICROSOFT MVP Receiving the Microsoft MVP award was one of the defining moments of Karen's career. She explains how years of consistently helping the community eventually resulted in recognition from Microsoft. More important than the title itself was the validation that her work was making a meaningful impact around the world.  STORIES THAT MADE THE BIGGEST IMPACT Among thousands of student success stories, Karen recalls one of the most unexpected: receiving a message from someone learning Excel while serving time in prison. She also shares emotional stories of people overcoming depression and rebuilding their confidence through learning. These experiences reinforced her belief that education is ultimately about empowering people—not simply teaching software.  QUICK FIRE FAVORITES Karen reveals some of her personal favorites during the rapid-fire round:Excel over Power BIBrigadeiro as her favorite Brazilian foodDark ModePower Query over Pivot TablesCtrl+K as her favorite shortcutCharts Maps as an underrated Excel featureWindows over MacClassroom training over online sessionsDashboard over reportsBackstreet Boys' As Long As You Love Me as her karaoke songOne word for Excel: FlexibilityBALANCING BUSINESS, FAMILY, AND CONTENT CREATION Karen closes the conversation by explaining that she doesn't separate work from life. Ideas flow naturally between family, teaching, entrepreneurship, and content creation. Rather than measuring success by hours worked, she focuses on making every hour meaningful while keeping her children at the center of everything she does. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.

    From Excel Expert to Microsoft MVP: Empowering Millions with Data, Dashboards & AI with Karen Abecia [Microsoft MVP]
  7. 3일 전

    The Copilot Credit Trap- Why Your AI Economy is Already Broken

    For decades, enterprise software followed a predictable financial model. Organizations purchased licenses, assigned them to users, and budgeted annual IT spending with confidence. AI changes that completely. Modern AI platforms are no longer sold purely as software—they're becoming consumption-based services where autonomous agents perform work on your behalf. Every action, every reasoning cycle, every orchestration task, and every AI workflow consumes credits instead of simply using a fixed license. This episode explains why Copilot Credits fundamentally change enterprise budgeting, why governance becomes more important than licensing, and how organizations must rethink identity, permissions, auditing, FinOps, and AI compliance before autonomous agents become part of everyday business operations. FROM SOFTWARE LICENSES TO AI ECONOMICS Traditional enterprise software was easy to budget. Organizations counted employees, purchased licenses, and forecasted annual costs with relatively little uncertainty. AI introduces a completely different financial model. Instead of paying only for access, organizations increasingly pay for work performed. Every autonomous action performed by an AI agent consumes credits based on: Reasoning complexityRuntimeContext sizeTool usageModel selectionThis transforms AI from a predictable software expense into an operational resource similar to cloud compute. The presentation argues that organizations are no longer purchasing software—they're purchasing autonomous labor, and that fundamentally changes IT economics. THE COPILOT CREDIT TRAP The biggest misconception surrounding Copilot Credits is that they simply represent another licensing model. They don't. Credits become the currency of AI work. A lightweight task may consume relatively few credits. Complex reasoning tasks involving multiple enterprise systems, long context windows, and autonomous orchestration consume dramatically more. Costs now scale according to: Agent behaviorTask complexityOrganizational adoptionWorkflow automationrather than simply employee count. Organizations may believe they have predictable AI costs because licensing appears fixed, while actual consumption grows continuously behind the scenes. This hidden variability creates what the presentation describes as the Copilot Credit Trap. WHY FINANCE CAN NO LONGER PREDICT COSTS Finance departments have traditionally planned annual software budgets using fixed subscription pricing. Consumption-based AI disrupts that model. Instead of budgeting for employees, organizations must now forecast: Daily agent activityDepartmental usageBusiness workflowsCredit consumptionSeasonal demandAutomation growthSmall changes in adoption can produce disproportionately large cost increases. The challenge isn't simply higher spending. It's the loss of financial predictability. Variable AI consumption introduces volatility that traditional IT budgeting processes were never designed to manage. VISIBILITY IS THE FIRST GOVERNANCE PROBLEM Many organizations cannot accurately answer basic questions such as: Which AI agents currently exist?Which departments deployed them?Which systems can they access?Which business processes do they automate?How much do they cost?The presentation describes this as the visibility crisis. Shadow AI deployments appear through: Copilot StudioPower AutomateDepartmental automationThird-party AI integrationsCustom workflowsWithout a complete inventory, governance becomes impossible because organizations cannot secure, monitor, or budget for systems they don't even know exist. PERMISSIONS BECOME MULTIPLIED One of the most significant risks discussed throughout the session is permission amplification. AI agents inherit the permissions of the identities under which they operate. If a user can access HR records, the agent can also access them. If a user can modify SharePoint documents, schedule meetings, or send emails, so can the agent. Unlike humans, however, agents perform these actions at machine speed and enterprise scale. This dramatically amplifies existing governance weaknesses, especially in environments suffering from years of permission creep and excessive data sharing. The presentation argues that AI doesn't create governance problems—it magnifies the ones organizations already have. AUTONOMY REQUIRES NEW GOVERNANCE Traditional software waits for users. Autonomous agents do not. Modern AI systems: Send emailsUpdate recordsSchedule meetingsTrigger workflowsCoordinate with other agentsoften after only an initial approval. As conditions change during execution, agents adapt automatically. This makes traditional approval processes insufficient. Organizations must introduce: Human approval gatesEscalation rulesSpending thresholdsRisk classificationsContinuous monitoringGovernance moves from documentation into active operational control. THE EU AI ACT CHANGES EVERYTHING One of the central themes of the presentation is the approaching regulatory landscape. Organizations deploying AI into HR, finance, customer services, or other sensitive business functions face increasing governance obligations under the EU AI Act. High-risk AI systems require: Risk managementTechnical documentationHuman oversightAudit trailsIncident reportingContinuous monitoringCompliance is no longer simply about technology. It becomes an enterprise operating capability involving legal, compliance, security, and business leadership working together. IDENTITY IS THE FOUNDATION The presentation argues that autonomous agents require independent identities rather than sharing user accounts. Each agent should receive: Dedicated identityScoped permissionsLeast-privilege accessIndependent audit trailLifecycle managementThis enables organizations to distinguish human actions from autonomous agent behavior while improving accountability and reducing operational risk. Identity becomes the foundation upon which every other governance capability depends Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.

    The Copilot Credit Trap- Why Your AI Economy is Already Broken
  8. 3일 전

    The End of AI Bloat: Why Modern Agents Need Skills

    Many AI agents start out fast, responsive, and surprisingly intelligent. But after a few months of real-world use, something changes. Response times increase, costs rise, prompts become enormous, and accuracy begins to decline. Organizations often respond by upgrading to larger models, expanding prompts, or adding more orchestration—but the underlying problem remains. The issue isn't the model. It's the architecture. This episode explains why monolithic prompts create what is known as the Context Tax, how modular Skills solve the problem through progressive disclosure, and why Skills are becoming the architectural foundation of modern AI agents across Microsoft Copilot Studio, GitHub Copilot, Claude Code, and the broader enterprise AI ecosystem. THE CONTEXT TAX Every enterprise AI project eventually faces the same challenge. At first, an agent contains a relatively small system prompt describing its role, tone, business rules, and guardrails. As the organization grows, more instructions are added: PoliciesCompliance rulesBusiness proceduresExamplesEdge casesDepartment-specific workflowsEventually the prompt becomes thousands of tokens long. Every user request forces the model to process every instruction—even when ninety-five percent of them are completely irrelevant. This hidden processing overhead is called the Context Tax. Rather than making agents smarter, larger prompts increase latency, raise inference costs, introduce reasoning noise, and gradually reduce answer quality. The presentation argues that the real problem isn't insufficient AI capability—it is forcing the model to continuously reason over information it doesn't actually need. WHY AGENTS DEGRADE OVER TIME Agent degradation is remarkably predictable. Organizations usually begin with one comprehensive instruction document that contains everything the AI should know. Initially this works well. Then new departments request additional functionality. Policies evolve. Compliance requirements expand. New workflows are added. Instead of restructuring the architecture, teams simply keep extending the same prompt. The result is context saturation. The model spends increasing amounts of effort searching through irrelevant guidance before finding the instructions that actually matter. This produces several side effects: Higher token consumptionSlower responsesIncreased hallucinationsMore inconsistent reasoningHigher operational costsThe AI hasn't become less intelligent. Its reasoning path has simply become overwhelmed by unnecessary context. ALWAYS-ON GUIDANCE VS SITUATIONAL EXPERTISE One of the most important architectural distinctions introduced in this session is separating always-on guidance from situational expertise. Always-on guidance includes information that applies to every conversation: Agent identityTone of voiceUniversal compliance rulesSecurity requirementsCore behavioral instructionsSituational expertise is different. It only matters when specific scenarios occur. Examples include: Vendor onboardingLeave eligibilityTax regulationsRefund workflowsRegional complianceIncident response proceduresTraditional agents mix both categories into one enormous prompt. Modern agent architectures separate them. Only universal guidance remains permanently loaded. Everything else becomes modular Skills that activate only when required. WHAT IS A SKILL? A Skill is much more than a prompt. It is a reusable package containing: Structured instructionsMetadataTrigger descriptionsOptional scriptsReference documentsTemplatesSupporting assetsThe core of every Skill is the SKILL.md file. This file defines: NameDescriptionPurposeTrigger conditionsWorkflowProcedural guidanceThe orchestrator doesn't initially load the entire Skill. Instead, it evaluates only the metadata. When the user's request matches the Skill description, the complete instructions are loaded into context. This dramatically reduces unnecessary reasoning while keeping specialist knowledge available exactly when needed. THE REASONING BOUNDARY The presentation introduces another important architectural concept: Skills define reasoning boundaries. Rather than forcing an AI model to treat every instruction as universally relevant, Skills establish clear expertise domains. A leave management Skill applies only to leave requests. A procurement Skill activates only during purchasing workflows. A compliance Skill loads only when compliance questions arise. Each Skill becomes an isolated reasoning domain. Instead of thinking about every possible business process simultaneously, the model focuses exclusively on the knowledge required for the current task. This improves both precision and consistency.  PROGRESSIVE DISCLOSURE One of the core design principles behind Skills is Progressive Disclosure. Instead of loading every instruction at startup, agents maintain only a lightweight catalog containing Skill names and descriptions. When a matching scenario appears: The orchestrator identifies the relevant Skill.The Skill loads into context.The task executes.The Skill unloads after completion.Everything else remains outside the context window. This significantly reduces: Token usageLatencyCompute requirementsInfrastructure costsThe architecture keeps the default state intentionally lean. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.

    The End of AI Bloat: Why Modern Agents Need Skills

평가 및 리뷰

5
최고 5점
3개의 평가

소개

Welcome to the M365.FM — your essential podcast for everything Microsoft 365, Azure, and beyond. Join us as we explore the latest developments across Power BI, Power Platform, Microsoft Teams, Viva, Fabric, Purview, Security, and the entire Microsoft ecosystem. Each episode delivers expert insights, real-world use cases, best practices, and interviews with industry leaders to help you stay ahead in the fast-moving world of cloud, collaboration, and data innovation. Whether you're an IT professional, business leader, developer, or data enthusiast, the M365.FM brings the knowledge, trends, and strategies you need to thrive in the modern digital workplace. Tune in, level up, and make the most of everything Microsoft has to offer. M365.FM is part of the M365-Show Network. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support.