Send us Fan Mail

Steve Corey looks ahead to Microsoft Ignite while examining the operational work required before and after an AI rollout. He explains why governance, oversharing, shadow AI, and endpoint protection remain ongoing concerns, then discusses how leaders should interpret major conference announcements that may change before general availability. The episode also explores expected developments around SharePoint, Autopilot, and Copilot Managed Runtime, including their potential implications for application development, licensing, security, and technical debt.

Topics Covered

• Preparing lightning talks and a table talk for Microsoft Ignite
• Data governance and oversharing during AI adoption
• Endpoint protection and shadow AI
• The risks of personal AI subscriptions in business environments
• The difference between announcements, previews, and general availability
• Expected developments involving SharePoint
• Hopes for Autopilot availability
• The purpose of Copilot Managed Runtime
• AI-assisted application development and “vibe coding”
• Licensing costs, security review, and technical debt

Key Takeaways

• Rolling out AI licenses begins an ongoing governance process rather than completing it.
• Conference announcements may change considerably before becoming production-ready.
• Rapid AI app creation should still include stability checks, security review, and red-team testing.
• Businesses should compare AI-built applications with current development effort, operating costs, and licensing requirements.
• Live conversations and collaboration can provide significant value at industry conferences.

Chapters

00:00 Welcome to the Briefing
00:37 Preparing for Ignite
01:28 Why the Vegas Trip Mattered
01:47 AI Rollout Risks
02:36 Shadow AI After Deployment
03:15 Personal AI and Data Protection
04:04 Oversharing and Information Governance
04:42 The Job Is Not Done
06:01 What to Expect at Ignite
06:33 Managing Announcement Anxiety
07:33 Announcements Versus General Availability
08:25 Lessons from Agent 365
09:36 Setting Expectations for Ignite
10:41 SharePoint Predictions
12:01 Hopes for Autopilot
12:56 The Value of Conference Conversations
13:56 Copilot Managed Runtime
14:25 From Pro Code to Low Code
16:25 AI-Built Applications
18:24 Security, Speed, and Red-Teaming
19:14 The Future of Low-Code Apps
21:03 Cost and Technical Debt
21:37 Closing Thought

⌚ Book your free on 1 on 1 call with me to discuss your business challenges and how AI can help 👉 https://go.stevecorey.com/copilotbriefing

🔴 Need help with Microsoft 365 Copilot adoption, governance, and employee training? Let's build your strategy ➡️ https://go.stevecorey.com/copilotbriefing

Send us Fan Mail

Microsoft is introducing more role-specific Copilot experiences, including a sales development agent designed to support outreach and prospect management. That potential comes with important questions about cost, grounding, voice, and autonomous communication.

The main discussion focuses on a common AI implementation mistake: picking the product before understanding the problem. Copilot Studio, custom-coded agents, Microsoft Foundry, and other approaches each have a place, but none should be treated as the automatic answer. The better method is to assess value, friction, complexity, maintainability, and future growth, then select the simplest solution that properly addresses the need.

Topics Covered

• Microsoft’s role-specific Copilot agents
• Sales outreach and prospect-management agents
• Grounding and voice in autonomous communication
• Copilot credits and potential operating costs
• Preparing a lightning talk for Microsoft Ignite
• Why Copilot Studio is not a universal entry point
• The range of agent-building options
• Risks of prematurely choosing custom AI
• High-value, low-complexity use cases
• Customer maintenance and self-sufficiency
• Right-sized solutions and upgrade paths

Key Takeaways

• Begin with requirements and desired business value, not a preferred AI platform.
• Compare complexity against the value and friction associated with the problem.
• Use Copilot Studio only when it fits the requirements.
• Verify how autonomous agents are grounded before they communicate externally.
• Avoid custom AI when a simpler Copilot or non-agent solution will work.
• Deliver solutions customers can maintain after implementation.
• Leave room to expand without overengineering the initial deployment.

Links

Sales Development Agent - https://learn.microsoft.com/en-us/dynamics365/sales/sales-dev-agent/sales-development-agent-overview

Lets meet up at Ignite - https://stevecorey.com/ignite

Need a Copilot solution? https://go.stevecorey.com/copilotbriefing

🔴 Need help with Microsoft 365 Copilot adoption, governance, and employee training? Let's build your strategy ➡️ https://go.stevecorey.com/copilotbriefing

Send us Fan Mail

Steve Corey examines why Microsoft 365 readiness is a foundational part of Copilot success. The discussion covers Copilot Code, SharePoint knowledge bases, cost controls, authoritative content, duplicate files, excessive access, abandoned workspaces, and inconsistent collaboration practices. Rather than waiting for a perfectly clean tenant, leaders should focus readiness work on defined AI use cases, high-value workflows, sensitive information, and measurable business outcomes.

❓ Need expert guidance with Microsoft 365 Copilot and agents - Let's talk about how I can help your business move forward - https://go.stevecorey.com/copilotbriefing-s1e6

Topics Covered

• Copilot Code and managed development environments
• Copilot feature and cost management
• SharePoint knowledge bases and Copilot credits
• Tenant context and its effect on AI results
• Authoritative SharePoint sites and sources of truth
• Duplicate files and outdated document versions
• Oversharing and permission risks
• Abandoned Teams and SharePoint sites
• Repeatable collaboration and knowledge-management patterns
• A prioritized Microsoft 365 readiness assessment

Key Takeaways

• Microsoft 365 readiness should be managed as a strategic business capability.
• Strong AI capabilities can still produce weak results in a disorganized information environment.
• High-value knowledge needs a clearly identifiable, maintained, authoritative location.
• Copilot surfaces existing access problems rather than creating them.
• Durable knowledge should be moved out of temporary collaboration channels.
• Important content and workspaces need accountable owners.
• Readiness work should begin with high-value workflows and sensitive content.
• AI-related change should be measured and reported against business outcomes.

⌚ Book your free on 1 on 1 call with me to discuss your business challenges and how AI can help 👉 https://go.stevecorey.com/copilotbriefing

🔴 Need help with Microsoft 365 Copilot adoption, governance, and employee training? Let's build your strategy ➡️ https://go.stevecorey.com/copilotbriefing