🎙️ Three MVPs at the table, a Copilot button that nobody uses, and a token limit that’s already maxed out by ten in the morning. In this episode, Robbert joins us again, and you can tell immediately. There’s laughter, teasing, and above all a lot of substance about where Copilot, Cowork, and Agent 365 really stand today.

We dive into statements about licenses, costs, and why “GA” by no means always means that everyone understands what it does. The EU data boundary comes up too, including the nuance that legally it’s all a lot less black-and-white than is often claimed. We also look at how choices around models, auto modes, and “critique” make not just a functional difference, but a financial one as well.

Robbert explains how frustration with tokens led to a Purview capability map, Danny shares concrete Cowork use cases with Power BI and Excel, and Hakim discovers how far you can get with a vague prompt like “I’m going on vacation, help me.” Meanwhile, plugins, skills, connectors, and federated Copilot connectors fly past your ears.

The episode ends more philosophically than planned. About always being on, notifications, focus blocks, social media, and why AI is actually interesting for us as a way to bring calm back into work that never seems to end. Less noise, more thinking. And yes, sometimes just turn Teams off and Copilot on.

The Copilot & AI at Work podcast is also available on Apple Podcasts, Podimo, Overcast, or in your other preferred podcast app via RSS.

🎙️ In this episode, we kick things off with a bit of self‑mockery and some good news. MVP announcements, reflecting on what we’ve actually done over the past year, and how something that started as “just an hour of chatting” has turned out to have quite a bit of impact.

We then quickly move into the content. This episode is intentionally a grab bag. We walk through everything that caught our attention in Copilot land over the past few weeks. From new agent‑like features in Outlook to calendars that increasingly manage themselves, and the question of when AI will really start doing things for you instead of just giving answers.

We spend quite some time on Cowork and why it feels different to us compared to many earlier AI features. Less steering, less micromanaging, and more the feeling of working with a colleague. At the same time, we also discuss the other side: model fatigue, benchmarks that don’t say much, automatic upgrades, and the question of what you actually need to keep up with and what you can safely ignore.

We also talk about Copilot Notebooks, specialised agents like the Legal Agent in Word, real‑time voice agents in Copilot Studio, and the idea that we’re slowly moving towards a world where UI for AI becomes more important than classic SaaS.

Along the way, we touch on PowerPoint surprises, vibe coding, security realities, LinkedIn irritations, and the upside of working in a time where no one really knows what this will all look like in a few years.

The Copilot & AI at Work podcast is also available on Apple Podcasts, Podimo, Overcast, or in your other preferred podcast app via RSS.

🎙️ A special episode, because for the first time we welcome a guest in our Dutch podcast. And not just any guest: Ivan Jankovic, Solutions Engineer at Microsoft, joins us at the table, immediately leading to a conversation that dives deep into the technology while always staying grounded in real-world practice.

We talk about Agent 365, agentic AI, governance, and security, and why these topics are becoming painfully relevant right now. Why is AI no longer just an IT party, but also not simply a toy for the business? And why does Copilot mainly expose issues that organizations have failed to get under control for years?

Ivan shares how he personally keeps up in a world where releases come faster than you can track them, why “AI first” often means something very different from what people expect, and why starting small and continuously iterating matters more than having a perfect strategy. We also discuss how agents will soon get their own identities, what that means for security and compliance, and why governance should not be a showstopper but something that runs in parallel.

Along the way, we hear real-world stories, from brute-force troubleshooting to bizarre customer locations, and why this wave of innovation is making even seasoned tech professionals excited to tinker again.

An episode about realism in the AI hype, the tension between business and IT, and why the future of work above all requires better questions, not just better tools.

The Copilot & AI at Work podcast is also available on Apple Podcasts, Podimo, Overcast, or in your other preferred podcast app via RSS.

For non-Dutch speaking subscribers we have added English subtitles to the video on YouTube if you are interested to watch the interview.