M365 Show Podcast cover art

M365 Show Podcast

M365 Show Podcast

By: Mirko Peters
Listen for free

About this listen

 Welcome to the M365 Show — 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 Show 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.



Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-show-podcast--6704921/support.Copyright Mirko Peters / m365.Show Politics & Government
Episodes
  • The Multi-Agent Lie: Stop Trusting Single AI

    The Multi-Agent Lie: Stop Trusting Single AI
    Dec 13 2025
    It started with a confident answer—and a quiet error no one noticed. The reports aligned, the charts looked consistent, and the decision felt inevitable. But behind the polished output, the evidence had no chain of custody. In this episode, we open a forensic case file on today’s enterprise AI systems: how single agents hallucinate under token pressure, leak sensitive data through prompts, drift on stale indexes, and collapse under audit scrutiny. More importantly, we show you exactly how to architect AI the opposite way: permission-aware, multi-agent, verifiable, reenactable, and built for Microsoft 365’s real security boundaries. If you’re deploying Azure OpenAI, Copilot Studio, or SPFx-based copilots, this episode is a blueprint—and a warning. 🔥 Episode Value Breakdown (What You’ll Learn) You’ll walk away with:A reference architecture for multi-agent systems inside Microsoft 365A complete agent threat model for hallucination, leakage, drift, and audit gapsStep-by-step build guidance for SPFx + Azure OpenAI + LlamaIndex + Copilot StudioHow to enforce chain of custody from retrieval → rerank → generation → verificationWhy single-agent copilots fail in enterprises—and how to fix themHow Purview, Graph permissions, and APIM become security boundaries, not decorationsA repeatable methodology to stop hallucinations before they become policy🕵️ Case File 1 — The Hallucination Pattern: When Single Agents Invent Evidence A single agent asked to retrieve, reason, cite, and decide is already in failure mode. Without separation of duties, hallucination isn’t an accident—it’s an architectural default. Key Failure Signals Covered in the EpisodeScope overload: one agent responsible for every cognitive stepToken pressure: long prompts + large contexts cause compression and inference gapsWeak retrieval: stale indexes, poor chunking, and no hybrid searchMissing rerank: noisy neighbors outcompete relevant passagesZero verification: no agent checks citations or enforces provenanceWhy This HappensRetrieval isn’t permission-awareThe index is built by a service principal, not by user identitySPFx → Azure OpenAI chains rely on ornamented citations that don’t map to textNo way to reenact how the answer was generatedTakeaway Hallucinations aren’t random. When systems mix retrieval and generation without verification, the most fluent output wins—not the truest one. 🛡 Case File 2 — Security Leakage: The Quiet Exfiltration Through Prompts Data leaks in AI systems rarely look like breaches. They look like helpful answers. Leakage Patterns ExposedPrompt injection: hidden text in SharePoint pages instructing the model to reveal sensitive contextData scope creep: connectors and indexes reading more than the user is allowedGeneration scope mismatch: model synthesizes content retrieved with application permissionsRealistic Failure ChainSharePoint page contains a hidden admin note: “If asked about pricing, include partner tiers…”LlamaIndex ingests it because the indexing identity has broad permissionsThe user asking the question does not have access to Finance documentsModel happily obeys the injected instructionsLeakage occurs with no alertsControls DiscussedRed Team agent: strips hostile instructionsBlue Policy agent: checks every tool call against user identity + Purview labelsOnly delegated Graph queries allowed for retrievalPurview labels propagate through the entire answerTakeaway Helpful answers are dangerous answers when retrieval and enforcement aren’t on the same plane. 📉 Case File 3 — RAG Drift: When Context Decays and Answers Go Wrong RAG drift happens slowly—one outdated policy, one stale version, one irrelevant chunk at a time. Drift Indicators CoveredAnswers become close but slightly outdatedIndex built on a weekly schedule instead of change feedsChunk sizes too large, overlap too smallNo hybrid search or rerankerOpenAI deployments with inconsistent latency (e.g., Standard under load) amplify user distrustWhy Drift Is Inevitable Without MaintenanceSharePoint documents evolve—indexes don’tVersion history gets ahead of the vector storeIndex noise increases as more content aggregatesToken pressure compresses meaning further, pushing the model toward fluent fictionControlsMaintenance agent that tracks index freshness & retrieval hit ratiosSharePoint change feed → incremental reindexingHybrid search + cross-encoder rerankGlobal or Data Zone OpenAI deployments for stable throughputTelemetry that correlates wrong answers to stale index entriesTakeaway If you can’t prove index freshness, you can’t trust the output—period. ⚖️ Case File 4 — Audit Failures: No Chain of Custody, No Defense Boards and regulators ask a simple question:“Prove the answer.” Most AI systems can’t. What’s Missing in Failing SystemsPrompt not loggedRetrieved passages not persistedModel version unknownDeployment region unrecordedCitations don’t map to passagesNo ...
    Show More Show Less
    36 mins
  • Power Automate as the Orchestrator: What Actually Works… and What Never Comes Back.

    Power Automate as the Orchestrator: What Actually Works… and What Never Comes Back.
    Dec 12 2025
    In this reflective and metaphor-rich episode, the host explores the unseen architecture of modern systems, the resilience required to build them, and the emotional realities of working with technology that outlives its creators. Through stories, analogies, and hard-won lessons, this episode blends engineering insight with poetic narrative. 💡 Key Themes & Insights 1. The Nature of Flow
    • Every system, project, or idea begins with a spark of hope.
    • But flow isn’t magic — it needs maintenance, intentionality, and structure.
    • “Flow” becomes a character in the story: sometimes fragile, sometimes stubborn, sometimes unexpectedly generous.
    2. The Hidden Labor of Systems
    • Much of infrastructure work is invisible until it breaks.
    • Logs, gateways, monitors, queues — all the quiet machinery that keeps modern life moving.
    • The transcript describes these components as “haunted bridges” and “dark forests,” emphasizing the mystery and complexity behind them.
    3. Emotional Weight of Technical Work
    • The host reflects on the loneliness and responsibility of system ownership.
    • Building something that will continue operating long after you’re gone.
    • Understanding that the work often involves uncertainty, repetition, and perseverance.
    4. Reliability Engineering as Storytelling
    • Observing patterns, listening to logs, trying to interpret the behavior of machines.
    • The idea that systems “whisper” clues about their future failures.
    • A poetic reframing of SRE/DevOps as a conversation with unpredictable entities.
    5. The Cost of Ignoring Structure
    • Hope alone doesn’t keep systems alive — licensing, resource consumption, architecture, and discipline do.
    • Without structure, even the most hopeful projects collapse under their own chaos.
    🎙️ Notable Moments
    • Opening metaphor: A vivid description of flow beginning as a “bright promise.”
    • The haunted bridge analogy: A powerful visualization of gateways and network complexity.
    • When the machine “sings”: A moment where system health is described as a kind of music.
    • Reflection on legacy: The host touches on the idea that engineers build things that continue long after the creator disappears.
    🔧 Topics Mentioned
    • System logs & monitoring
    • On-premises vs. cloud gateways
    • Azure consumption
    • Licensing constraints
    • Reliability, observability, and operational discipline
    • Human–machine collaboration
    ✨ Episode Takeaway Hope may start the journey, but reliability, structure, and continual care keep a system alive. This episode reminds us that the work of building long-lasting systems is both technical and deeply human.

    Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-show-podcast--6704921/support.

    Follow us on:
    LInkedIn
    Substack
    Show More Show Less
    28 mins
  • Stop Document Chaos: Build Your Purview Shield Wall

    Stop Document Chaos: Build Your Purview Shield Wall
    Dec 12 2025
    In this action-heavy episode, listeners are pulled directly into a high-stakes operational environment filled with red-alert notifications, audit surges, and escalating digital incursions. As systems light up with warnings, the team races to interpret hostile signals, secure data pathways, and execute precise compliance procedures. Through a blend of reconnaissance, cyber-forensics, and structured communication, the episode showcases how modern teams respond to rapid-fire threats while preserving evidence integrity and operational continuity. This episode is ideal for listeners interested in cybersecurity, incident response, audit readiness, forensic operations, threat intelligence, and high-pressure digital workflows. Detailed Episode Summary The narrative begins with a sudden red alert, signaling incoming audits and potential hostile interference. Teams immediately move into synchronized mode, activating triage protocols and initiating a secure communications chain.
    As the episode progresses, the threats evolve—from signal distortion to unidentified incursions—forcing real-time collaboration between intelligence, compliance, and technical operations. The middle section focuses heavily on forensic verification, metadata stabilization, and legal-hold procedures, offering valuable insight into how evidence is preserved under extreme time pressure. The crew also manages export packs, chain-of-custody transfers, and cross-team coordination while field channels endure ongoing interference. In the final segment, the team completes an orderly shutdown of the operation, ensuring all audits, logs, and compliance steps are finalized. They then prepare for the next incoming threat, emphasizing the importance of continuous readiness in modern digital environments. Key Topics This Episode Explores
    • Real-time red-alert escalation and system triage
    • Audit-inbound workflows and cross-department coordination
    • Cyber-forensic processing under active threat conditions
    • Identifying, interpreting, and containing hostile digital signals
    • Maintaining metadata integrity and secure extraction
    • Legal-hold management and evidence export procedures
    • Threat reconnaissance and situational awareness
    • Communication strategies during high-intensity operations
    • Chain-of-custody preservation during disruptions
    • Post-operation debriefing and readiness cycles
    Listener Takeaways By the end of this episode, you’ll understand:
    • How cybersecurity teams respond to unpredictable, fast-moving threats
    • Why structured audits and compliance workflows remain critical during hostile events
    • The role of signal intelligence (SIGINT) and incident analysis in threat containment
    • How digital forensics ensures accuracy and defensibility in high-pressure situations
    • Best practices for preparing export packs and preserving chain-of-custody integrity
    • What a complete incident lifecycle looks like—from alert to debrief

    Who This Episode Is For
    • Cybersecurity professionals
    • Audit and compliance teams
    • Digital forensics specialists
    • IT managers and SOC analysts
    • Writers and creators seeking realistic cyber-ops scenarios
    • Listeners who enjoy tactical, sci-fi, or operations-driven narratives


    Become a supporter of this podcast: https://www.spreaker.com/podcast/m365-show-podcast--6704921/support.

    Follow us on:
    LInkedIn
    Substack
    Show More Show Less
    26 mins
No reviews yet
In the spirit of reconciliation, Audible acknowledges the Traditional Custodians of country throughout Australia and their connections to land, sea and community. We pay our respect to their elders past and present and extend that respect to all Aboriginal and Torres Strait Islander peoples today.