Your AI Agents Are Orphaned- The Structural Shift to Agent ID

Your AI Agents Are Orphaned- The Structural Shift to Agent ID

Artificial intelligence is changing enterprise identity faster than most organizations realize. Every week, new AI agents are being deployed across Microsoft 365 tenants, accessing SharePoint, Microsoft Graph, Teams, Exchange, and business applications. Yet in many organizations, nobody can confidently answer one simple question: Who actually owns these agents? In this episode, we explore why traditional Service Principals were never designed for autonomous AI systems and why Microsoft introduced Entra Agent ID as an entirely new identity model. You'll learn how AI governance is shifting from application management toward identity-first governance, where every agent becomes a managed digital worker with accountability, lifecycle management, and built-in security. WHY AI AGENTS HAVE BECOME A GOVERNANCE CHALLENGE Many organizations already have AI agents running inside their Microsoft 365 environment without realizing how difficult they are to govern. Copilot Studio bots, automation workflows, custom Graph applications, and AI assistants often appear organically across departments. Projects finish, developers change roles, but the identities remain active, continuing to access corporate resources without a clearly defined owner. This creates a growing population of orphaned non-human identities. Traditional governance processes were designed around employees and applications—not autonomous systems capable of making decisions, calling tools, and accessing sensitive enterprise data. As organizations move toward thousands of AI agents, this architectural mismatch becomes increasingly difficult to manage.  WHY SERVICE PRINCIPALS NO LONGER SCALE The episode explains why Service Principals struggle to support modern AI workloads. They were originally designed for long-lived backend applications with predictable behavior, not dynamic AI agents that may exist for only minutes, collaborate with other agents, or require unique permissions for individual tasks. These limitations create several operational problems: * Credential sprawl across thousands of agents * Limited auditability * Shared identities that reduce visibility * No built-in ownership model * Difficult lifecycle management Instead of solving governance, Service Principals often become the source of governance complexity. MICROSOFT ENTRA AGENT ID Microsoft's response is Agent ID, a new identity type built specifically for autonomous AI systems. Rather than hiding agents behind application registrations, each AI agent becomes a first-class identity inside Microsoft Entra ID with its own lifecycle, audit trail, sponsorship, and governance controls. A major innovation is the concept of Blueprints. Instead of creating hundreds of individual identities manually, administrators define reusable templates that centrally manage authentication, permissions, and governance. Every AI agent inherits these controls while remaining individually traceable throughout its lifecycle. GOVERNANCE BECOMES PART OF THE ARCHITECTURE One of the most important ideas discussed in the episode is that governance should no longer depend on documentation or manual processes. Instead, governance becomes an architectural capability built directly into the identity platform. Sponsors, access reviews, lifecycle policies, Conditional Access, audit logging, and permission inheritance all become native characteristics of the identity itself rather than separate administrative tasks. This shift dramatically reduces orphaned identities while creating a complete chain of accountability from every AI action back to an identifiable human sponsor.  SECURITY, COMPLIANCE AND DATA PROTECTION AI agents increasingly operate across SharePoint, Teams, Exchange, OneDrive and line-of-business applications, making identity governance inseparable from data governance. The episode explains how Microsoft Entra Agent ID integrates with Microsoft Purview, Defender, Conditional Access and Identity Protection to extend Zero Trust principles to autonomous AI. Topics covered include: * Agent-specific Conditional Access * Identity Protection and behavioral baselines * Microsoft Purview integration * Data Loss Prevention (DLP) * Risk detection and runtime protection Together these capabilities allow organizations to apply the same governance standards to AI agents that already exist for human identities. BUILDING AN AGENTIC ENTERPRISE The episode also introduces a practical roadmap for adopting Agent ID. Rather than simply enabling a new Microsoft feature, organizations should begin by discovering existing AI agents, assigning business sponsors, creating reusable blueprints, grounding agents on trusted enterprise data, and gradually integrating governance into everyday operations. By combining identity management, security, compliance, and operational governance into a single architecture, Agent ID enables organizations to scale AI safely while maintaining visibility and accountability across every autonomous system they deploy. KEY TAKEAWAYS Agent ID is far more than another Microsoft security feature—it represents a fundamental shift in how enterprises manage AI. Organizations are moving away from treating AI agents as anonymous applications and toward managing them as governed digital workers with identities, sponsors, lifecycle management, security controls, and complete auditability. Those who adopt this model early will be significantly better positioned to scale enterprise AI securely while meeting future governance and compliance requirements. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support [https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support?utm_source=rss&utm_medium=rss&utm_campaign=rss].

The Architecture of Agility: Bicep at Scale

Modern cloud platforms don't fail because Azure isn't powerful enough—they fail because governance, automation, and developer experience weren't designed to scale together. In this episode of the M365 FM Podcast, host Mirko Peters explores how Azure Bicep evolves from a simple Infrastructure as Code language into the foundation of enterprise platform engineering. Rather than focusing on syntax, this deep dive examines the architectural principles that allow organizations to build Azure Landing Zones, governance models, subscription strategies, reusable modules, and self-service platforms that enable teams to innovate without sacrificing security or compliance. The episode challenges one of the biggest assumptions in enterprise IT: that tighter control automatically creates better governance. Instead, you'll discover why modern cloud platforms succeed by making the governed path the easiest path. From subscription vending and management groups to policy-as-code, FinOps, observability, and platform engineering, this episode provides a blueprint for designing Azure environments that remain agile as organizations grow. WHY LANDING ZONES ARE REALLY GOVERNANCE MODELS Azure Landing Zones are often presented as technical architectures, but they're much more than networking diagrams and subscription hierarchies. Every decision encoded in a Bicep module determines who can deploy infrastructure, who owns resources, how quickly teams can provision environments, and how governance flows across the organization. The episode explains why successful cloud platforms move away from gatekeeper governance toward enablement, replacing manual approvals with automated guardrails that empower teams while maintaining security and compliance.  BUILDING FOR AGILITY, NOT CONTROL Many enterprises unintentionally create Shadow IT by making official infrastructure too slow to consume. Instead of preventing risk, excessive governance often encourages developers to work around official processes. Topics include: * Subscription vending * Self-service infrastructure * Platform engineering * Developer experience * Shadow IT * Governance as Code * Automation * Azure Landing Zones * Organizational agility * Cloud operating models You'll learn why reducing friction often improves security more effectively than adding additional approval processes. SCALING AZURE WITH REUSABLE BICEP MODULES As organizations grow, infrastructure cannot rely on copy-and-paste templates. Reusable, versioned Bicep modules become the building blocks for enterprise platforms. The discussion explores: * Azure Verified Modules (AVM) * Module registries * Semantic versioning * Infrastructure contracts * Composition patterns * Parameter design * Shared variables * Resource modules * Solution modules * Version control Rather than maintaining hundreds of custom templates, organizations can publish standardized modules that continuously evolve while maintaining compatibility across teams. MANAGEMENT GROUPS, POLICY & PLATFORM GOVERNANCE Governance begins long before a resource is deployed. This episode explains how Management Groups, Azure Policy, RBAC, and Policy Initiatives combine to create scalable governance models that automatically enforce organizational standards. Instead of treating policies as deployment blockers, you'll discover how audit-first strategies, policy-as-code, and automated compliance enable organizations to maintain security without slowing down innovation.  PLATFORM ENGINEERING AND THE GOLDEN PATH The role of central IT is changing. Rather than acting as infrastructure gatekeepers, platform teams increasingly operate as internal product teams. The episode explores how self-service infrastructure, golden paths, reusable templates, and developer-first experiences allow application teams to provision secure Azure environments within minutes instead of waiting weeks for approvals. Platform engineering shifts the focus from enforcing permissions to enabling productivity.  FINOPS, OBSERVABILITY & COST GOVERNANCE Enterprise cloud success isn't measured solely by uptime. Organizations must also understand infrastructure costs, policy compliance, operational health, and governance effectiveness. Topics include: * FinOps * Cost governance * Azure Monitor * Log Analytics * Diagnostic Settings * Observability * Tagging strategies * Chargeback * Showback * Continuous compliance You'll learn how telemetry, policy enforcement, and financial accountability combine to create cloud environments that remain sustainable as Azure adoption grows. WHO SHOULD LISTEN? This episode is ideal for: * Azure Architects * Platform Engineers * DevOps Engineers * Cloud Engineers * Enterprise Architects * Infrastructure Engineers * Azure Administrators * Security Architects * IT Decision Makers * Microsoft MVPs * Anyone designing enterprise Azure platforms Whether you're building Azure Landing Zones, implementing Azure Bicep, creating reusable Infrastructure as Code modules, modernizing governance, or adopting Platform Engineering, this episode provides practical guidance for building cloud platforms that balance agility, security, automation, and scalability. If you want to understand how successful enterprises use Azure Bicep not just as a deployment language, but as the foundation for modern cloud governance and platform engineering, this episode offers a comprehensive roadmap for designing Azure environments that can scale with both technology and business growth. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support [https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support?utm_source=rss&utm_medium=rss&utm_campaign=rss].

Designing the Future. AI, UX, and the Next Generation of Microsoft Power Platform with Tchesco Ayih [MVP-MCT]

In this episode of the M365 Podcast, Mirko Peters sits down with Tchesco Ayih, Microsoft MVP, Microsoft Certified Trainer (MCT), international speaker, mentor, and Power Platform expert. Tchesco shares his inspiring journey into the Microsoft ecosystem, discusses the growing importance of user experience (UX) in low-code development, and explores how artificial intelligence is transforming the way organizations build business applications. From Power Apps and Dataverse to Copilot Studio and AI-powered development, this conversation provides valuable insights for anyone working with Microsoft technologies, citizen development, low-code platforms, or digital transformation initiatives. Tchesco explains why beautiful design is not just a nice-to-have feature but a critical component of successful business applications, and why user-centric thinking should always come before functionality. FROM TRADITIONAL DEVELOPMENT TO MICROSOFT POWER PLATFORM Tchesco's journey into the Microsoft ecosystem started as a traditional developer with a passion for user interfaces and design. After earning Azure certifications and sharing his learning journey online, an opportunity emerged that introduced him to Microsoft Power Platform. What initially began as curiosity quickly developed into a career dedicated to helping organizations build efficient business solutions using low-code technologies. Throughout the discussion, Tchesco highlights how continuous learning, community engagement, and technical curiosity played essential roles in shaping his professional path. His story serves as a reminder that modern technology careers often evolve through experimentation, networking, and a willingness to embrace emerging platforms.  THE MVP AND MCT JOURNEY Becoming a Microsoft MVP and Microsoft Certified Trainer did not happen overnight. Tchesco discusses how helping others in online communities became the foundation of his recognition within the Microsoft ecosystem. By answering technical questions, delivering community sessions, supporting Microsoft Learn Student Ambassadors, and consistently sharing knowledge, he built a reputation that eventually led to an MVP nomination. Key lessons from his journey include: * Consistently contribute to technical communities. * Share knowledge freely and help solve real problems. * Build expertise through hands-on experience. * Pursue Microsoft certifications to validate skills. * Seek mentorship from experienced professionals. * Give back to the community that helped your own growth. WHY UX IS THE SECRET INGREDIENT OF GREAT BUSINESS APPS One of the most passionate parts of the discussion centers around User Experience (UX). According to Tchesco, many Power Platform developers become overly focused on functionality and process automation while neglecting how users actually experience the application. A business process may work perfectly from a technical perspective, but if the application feels confusing, cluttered, or difficult to navigate, user adoption can suffer significantly. Tchesco argues that every application should be designed with accessibility, usability, and user satisfaction in mind. He emphasizes that developers should think beyond buttons and forms and consider: * User accessibility requirements. * Application navigation flows. * Visual hierarchy and layout. * User feedback cycles. * Readability and simplicity. * Long-term adoption and engagement. Rather than viewing UX as a separate discipline, Tchesco believes UX should be integrated into every stage of Power Platform development. DESIGN FIRST, FEATURES SECOND One of the most powerful takeaways from the conversation is Tchesco's belief that developers should begin with the user interface before implementing features. Many developers immediately start building databases, business logic, and automation workflows. However, Tchesco recommends designing the application's screens first, creating a clear structure for how users will interact with information before any technical implementation begins. A thoughtful interface helps developers: * Organize business processes logically. * Clarify data requirements. * Reduce development rework. * Improve stakeholder alignment. * Create more intuitive user journeys. * Increase overall user satisfaction. This design-first mindset helps teams build solutions that are both technically effective and enjoyable to use. UNDERSTANDING THE MICROSOFT POWER PLATFORM ECOSYSTEM For listeners new to the Power Platform, Tchesco provides a practical overview of the platform's major components and how they work together to accelerate digital transformation initiatives. The Microsoft Power Platform consists of: * Power Apps for low-code application development. * Power Automate for workflow automation. * Dataverse as a scalable relational data platform. * Power Pages for external-facing business websites. * Copilot Studio for conversational AI experiences. * Connectors that integrate hundreds of business systems. Together, these tools enable organizations to rapidly automate processes, modernize legacy workflows, and reduce development timelines that traditionally required months of custom coding. DATAVERSE: THE FOUNDATION OF MODERN BUSINESS SOLUTIONS The conversation explores why Dataverse has become one of the most important components of the Power Platform ecosystem. Tchesco describes Dataverse as much more than a simple database. Its relational architecture, deep integration with Microsoft services, security model, and performance capabilities make it a strong choice for enterprise-grade business applications. Compared to traditional approaches relying heavily on SharePoint lists, Dataverse enables more scalable and maintainable business solutions while supporting advanced application scenarios and automation workflows. AI, COPILOT STUDIO, AND THE FUTURE OF APPLICATION DEVELOPMENT Artificial Intelligence plays a major role throughout the discussion. Tchesco shares how AI has already transformed many aspects of his daily work, helping accelerate development, generate ideas, improve productivity, and reduce repetitive tasks. The conversation explores the growing role of: * AI-assisted development. * Prompt engineering. * Copilot Studio. * Agent-based experiences. * AI-powered business automation. * Conversational user interfaces. Rather than replacing developers, Tchesco views AI as a powerful assistant that increases productivity when used correctly. He compares AI to a junior colleague that can help complete tasks faster, while experienced professionals remain responsible for validation, governance, and final decision-making. PROMPT ENGINEERING IS BECOMING A CRITICAL SKILL One particularly interesting topic is the emergence of prompt engineering as an essential skill for modern developers. According to Tchesco, professionals who learn how to communicate effectively with AI systems will gain a significant advantage. Knowing how to provide clear context, detailed requirements, and accurate instructions can dramatically improve the quality of AI-generated outputs. As AI becomes increasingly embedded across Microsoft products and business applications, prompt-driven development is likely to become a standard part of the software development process.  GOVERNANCE AND RESPONSIBLE AI ADOPTION While enthusiastic about AI's potential, Tchesco also stresses the importance of governance. Organizations must establish clear guidelines around: * Data security. * Privacy protection. * AI usage policies. * Information sharing restrictions. * Compliance requirements. * Human oversight and validation. Successful AI adoption requires balancing innovation with responsible governance. Businesses that introduce AI without adequate controls may expose sensitive information or create risks that outweigh the benefits. COMMUNITY, LEARNING, AND GIVING BACK Beyond technology, this episode highlights the importance of community engagement. Tchesco attributes much of his professional growth to the Microsoft community. By participating in forums, speaking at conferences, mentoring others, and sharing knowledge, he has developed both technical expertise and professional opportunities. His message is clear: growth accelerates when individuals actively contribute rather than simply consume information. Whether through local user groups, virtual events, online forums, or international conferences, communities remain one of the most valuable resources available to technology professionals.  LOOKING AHEAD: THE FUTURE OF LOW-CODE DEVELOPMENT As the conversation concludes, Tchesco shares his perspective on the future of low-code technology. He believes low-code platforms will continue to expand rapidly over the coming years, enabling organizations to build solutions faster while empowering more people to participate in application development. At the same time, strong foundations in UX, governance, architecture, and problem-solving will remain essential. Technology may evolve, but understanding users and solving business challenges effectively will always be the real differentiator. For anyone interested in Microsoft Power Platform, AI-driven development, business applications, digital transformation, or user-centric design, this episode delivers practical insights and forward-looking perspectives from one of the community's most passionate advocates. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support [https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support?utm_source=rss&utm_medium=rss&utm_campaign=rss].

Azure Bicep Fundamentals for Real Projects

Learning Azure Bicep is easy. Building infrastructure that survives real-world enterprise environments is something entirely different. In this episode of the M365 FM Podcast, host Mirko Peters explores what separates hobbyist Infrastructure as Code from production-ready cloud architecture. Rather than focusing on syntax alone, this episode dives into the architectural thinking behind Azure Bicep, showing how modern organizations design scalable, secure, and governable Azure environments using Infrastructure as Code, Azure Policy, Deployment Stacks, Azure Verified Modules, Management Groups, and enterprise governance. You'll learn why Bicep is far more than a replacement for ARM Templates. It has become Microsoft's strategic language for defining Azure infrastructure, enforcing governance, automating deployments, and enabling cloud platforms that remain maintainable as organizations grow. Whether you're deploying your first storage account or designing enterprise landing zones across hundreds of Azure subscriptions, this episode provides the principles needed to build infrastructure that lasts. WHY BICEP IS MORE THAN BETTER SYNTAX Many developers view Bicep simply as an easier way to write ARM Templates. While the cleaner syntax, reusable modules, and improved readability are significant improvements, Microsoft's long-term vision goes much further. Bicep has evolved into the language that defines Azure governance, policy, security baselines, and infrastructure lifecycle management. Modern Infrastructure as Code is no longer about automating deployments—it's about codifying architectural decisions that remain consistent across teams, subscriptions, and cloud environments.  THINKING LIKE A CLOUD ARCHITECT One of the central themes throughout the episode is the difference between automation and architecture. Real-world Bicep projects are built around organizational standards rather than individual deployments. Topics include: * Infrastructure as Code * Enterprise architecture * Governance as Code * Management Groups * Azure Landing Zones * Resource ownership * Organizational standards * Platform engineering * Cloud operating models * Infrastructure lifecycle You'll discover why enterprise Bicep templates become long-term blueprints that survive team changes, organizational growth, and evolving business requirements. BUILDING REUSABLE BICEP MODULES As Azure environments become larger, reusable modules become the foundation of maintainable infrastructure. The episode explains how well-designed modules encapsulate organizational decisions including security baselines, naming conventions, monitoring requirements, compliance standards, and networking patterns. Rather than copying infrastructure between projects, organizations build internal module libraries that allow development teams to deploy secure infrastructure while automatically following corporate standards. The discussion covers: * Azure Verified Modules (AVM) * Resource modules * Pattern modules * Platform modules * Module registries * Versioning * Parameter validation * Variables * Outputs * Semantic versioning DEPLOYMENT STACKS AND INFRASTRUCTURE LIFECYCLE One of Microsoft's biggest Infrastructure as Code innovations is Deployment Stacks. Unlike traditional ARM deployments, Deployment Stacks introduce state awareness, allowing Azure to understand which resources belong together throughout their lifecycle. The episode explores how Deployment Stacks simplify environment management, resource cleanup, brownfield migrations, and governance while reducing orphaned resources and configuration drift. You'll also learn how Deployment Stacks compare with Terraform's state management model and why this represents a major architectural shift for Azure-native deployments.  MANAGEMENT GROUPS, POLICY, AND GOVERNANCE Enterprise Azure environments cannot be managed one subscription at a time. This episode explains why Management Groups form the true control plane for Azure governance. Topics include: * Azure Policy * Policy Initiatives * RBAC * Least Privilege * Management Group hierarchy * Subscription governance * Compliance * Security baselines * Governance as Code * Platform engineering By defining governance at the Management Group level, organizations can automatically enforce consistent security, compliance, and operational standards across hundreds of subscriptions. BUILDING ENTERPRISE LANDING ZONES Azure Landing Zones are much more than preconfigured subscriptions. They provide complete enterprise-ready cloud environments including networking, monitoring, identity integration, governance, security controls, and operational best practices. The discussion explores how organizations compose Landing Zones using reusable Bicep modules, allowing application teams to deploy workloads into standardized cloud environments without sacrificing flexibility.  BICEP VS TERRAFORM One of the most common questions facing cloud architects is whether to standardize on Azure Bicep or Terraform. Rather than declaring a winner, this episode explains the strengths of both platforms. Bicep provides immediate support for new Azure services, seamless Azure integration, and simplified Azure-native deployments. Terraform remains an excellent choice for organizations operating across Azure, AWS, Google Cloud, and hybrid environments. The right decision depends on business requirements, governance strategy, and long-term platform goals—not simply personal preference.  WHO SHOULD LISTEN? This episode is perfect for: * Azure Architects * Cloud Engineers * DevOps Engineers * Platform Engineers * Infrastructure Engineers * Azure Administrators * Enterprise Architects * Cloud Consultants * IT Decision Makers * Microsoft MVPs * Anyone learning Azure Bicep Whether you're writing your very first Bicep file, migrating from ARM Templates, implementing Azure Landing Zones, designing enterprise governance, or building reusable Infrastructure as Code libraries, this episode provides practical guidance drawn from real-world cloud architecture. If you want to move beyond simple deployments and start building Azure environments that are secure, scalable, reusable, and maintainable for years to come, this episode offers a comprehensive foundation for mastering Azure Bicep in real enterprise projects. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support [https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support?utm_source=rss&utm_medium=rss&utm_campaign=rss].

Beyond Prompt a Page: Building Enterprise-Ready AI Experiences in Power Apps with Sara Lagerquist [MVP]

Artificial Intelligence is changing how organizations build business applications, but moving beyond simple demos requires much more than adding a chatbot or generating code with AI. In this episode of the M365 Show, Microsoft MVP Sara Lagerquist shares how enterprise organizations can use Generative Pages, Model-Driven Apps, and the Microsoft Power Platform to build scalable, maintainable, and production-ready AI experiences. From architecture and governance to citizen development and modern application design, this conversation explores what the next generation of business applications looks like inside the Microsoft ecosystem. FROM BUSINESS ADMINISTRATOR TO MICROSOFT MVP Sara shares her unique journey into technology. Without a traditional IT background, she discovered Microsoft Dynamics CRM while working as an administrator and quickly realized that designing business solutions came naturally. Over the past twelve years she has evolved into an Applications Architect, Microsoft MVP, international speaker, and one of the leading experts in the Microsoft Power Platform. Her story demonstrates that understanding business processes can be just as valuable as writing code when building enterprise solutions.  WHY MODEL-DRIVEN APPS ARE STILL THE FOUNDATION While Canvas Apps receive significant attention, Sara explains why Model-Driven Apps continue to be her preferred foundation for many enterprise solutions. They provide built-in navigation, security, search capabilities, responsive layouts, and rapid development that dramatically reduces implementation time. Rather than replacing Model-Driven Apps, Microsoft is extending them with Generative Pages, allowing developers to create highly customized user experiences without sacrificing the benefits of the underlying platform. The result is a powerful combination of low-code productivity and enterprise scalability.  GENERATIVE PAGES: THE NEXT EVOLUTION OF POWER APPS One of the biggest topics in the episode is Microsoft's new Generative Pages feature. Sara explains why she believes Generative Pages represent the future of Power Apps by combining AI-assisted development with highly flexible user interfaces. Instead of spending days or weeks creating custom pages manually, developers can use AI to rapidly generate complex interfaces that integrate seamlessly into Model-Driven Apps. The discussion covers current capabilities, limitations, supported regions, preview features, and Microsoft's roadmap for expanding Generative Pages across the Power Platform.  WHY VISUAL STUDIO CODE OUTPERFORMS THE MAKER PORTAL Although Microsoft offers AI-powered page generation directly inside the Power Apps Maker Portal, Sara recommends a different workflow for professional development. She explains how Visual Studio Code, GitHub Copilot, and Microsoft's official Generative Pages extension provide significantly better results. Developers gain access to stronger AI models, planning workflows, wireframe generation, version control, and a far more efficient prompting experience than the browser-based interface currently provides. For organizations building production applications, this development approach offers greater flexibility while reducing costs and improving overall quality.  PROMPT ENGINEERING FOR ENTERPRISE APPLICATIONS Building AI-generated business applications starts long before writing the first prompt. Sara discusses how successful prompting begins with understanding business processes rather than focusing on user interface design. Instead of asking AI to simply create a page, architects should clearly define business goals, user journeys, data relationships, and expected outcomes. The conversation also explores iterative prompting, planning conversations with AI, wireframe generation, and using AI itself to improve prompt quality before development even begins.  DESIGNING FOR LONG-TERM MAINTAINABILITY One of the biggest challenges in enterprise development is ensuring applications remain understandable long after the original developer has moved on. Sara explains how AI can actually improve documentation by automatically generating architecture diagrams, implementation plans, process documentation, user manuals, screenshots, and development guidance throughout the entire project lifecycle. Rather than making applications harder to maintain, AI can significantly improve knowledge transfer when used correctly.  NEW CAPABILITIES COMING TO GENERATIVE PAGES During the interview, Sara shares several exciting announcements from the European Power Platform Conference. Upcoming improvements include support for additional data sources beyond Dataverse, enabling developers to integrate external systems directly into Generative Pages. Microsoft is also introducing the ability to embed Generative Pages directly inside Model-Driven App forms, opening entirely new possibilities for creating rich enterprise user experiences without relying on traditional custom development. These features have the potential to reshape how Power Apps solutions are designed over the coming years.  GOVERNANCE IN THE AGE OF AI As AI dramatically accelerates application development, governance becomes even more important. Sara discusses why organizations must shift their focus from controlling development speed to managing adoption, ownership, lifecycle management, security, and long-term maintenance. With AI capable of producing applications much faster than before, businesses need clear governance frameworks to ensure solutions remain secure, scalable, and aligned with business objectives. The conversation also explores Microsoft's evolving governance capabilities and the future of the Power Platform Admin Center.  THE FUTURE OF CITIZEN DEVELOPMENT Citizen developers continue to play an increasingly important role within modern organizations. Sara believes AI will lower technical barriers even further, allowing business users to build sophisticated solutions with minimal coding experience. However, she emphasizes that successful citizen development requires education around governance, security, architecture, and business process design rather than simply teaching users where to click inside Power Apps. The future belongs to professionals who understand both technology and business processes.  BUILDING BETTER BUSINESS PROCESSES One of the strongest messages throughout the episode is that organizations should stop thinking about recreating existing processes with AI. Instead, AI provides an opportunity to redesign how work gets done entirely. Rather than building "faster horses," businesses should rethink workflows from the ground up, creating entirely new experiences that take advantage of automation, intelligent decision-making, and modern low-code platforms. This shift in thinking may ultimately become the biggest transformation AI brings to enterprise software development.  COMMUNITY, LEARNING, AND STAYING CURRENT Sara also shares the story behind founding the Power Platform Community Sweden, which has grown into one of the country's most active communities for makers, consultants, and architects. She discusses the importance of conferences, community events, collaboration, and continuous learning in an ecosystem that evolves almost every month. With new AI features, Power Platform capabilities, and Microsoft announcements arriving constantly, staying connected to the community has become essential for every modern architect and developer.  RAPID FIRE INSIGHTS The episode concludes with a fun rapid-fire session covering Sara's favorite development habits, career advice, conference experiences, books, and lessons learned throughout her Microsoft journey. Her biggest takeaway for listeners is simple: explore Generative Pages today—but build them using professional development tools rather than relying solely on the Maker Portal. For anyone interested in enterprise AI development with Microsoft Power Apps, this episode provides an excellent starting point. Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support [https://www.spreaker.com/podcast/m365-fm-modern-work-security-and-productivity-with-microsoft-365--6704921/support?utm_source=rss&utm_medium=rss&utm_campaign=rss].