Welcome
Check in, settle in and meet the Midlands Microsoft community.
Midlands
Photo by Lulu Black on Unsplash
Photo by Lulu Black on Unsplash
Microsoft products and technology
All experience levels
New speakers welcome
We use a single-room format, with sessions covering a range of Microsoft products and technology.
We are open to people who work with, are learning about or are interested in Power Platform, Azure, Microsoft 365, Dynamics, AI, Copilot, data, security and related topics.
Times are approximate and may shift slightly on the day.
Check in, settle in and meet the Midlands Microsoft community.
Dieter Gobeyn
Most agent demos skip the question: what does an agent actually cost to run? A REST call costs the same every time. An agent doesn't, since the same prompt varies with reasoning depth, tool calls, retries and context growth. We'll map what you really pay for in Microsoft Foundry beyond model tokens and then look at several architectural options for controlling spend.. plus when NOT to build an agent at all. Expect some live demos in AI Foundry portal, and you'll leave with a FinOps checklist.
Five minutes.
Arjan Rijsdijk
You’ve probably heard the phrase “Low-code is dead — as we know it” more than once. But is that really true? Is low-code truly disappearing, or is it simply evolving? The Power Platform is changing rapidly under the influence of AI. Building apps is no longer just about dragging controls onto a canvas or writing formulas. It’s about expressing intent and letting agents do the heavy lifting. AI-driven “vibe coding” promises speed, flexibility, and a new way of working. But what does that look like in practice? In this session, we take a realistic look at the vibe coding capabilities of the Power Platform. Through hands-on demos, you’ll see how to create generative pages for model-driven apps, build applications via vibe.powerapps.com and develop Code Apps using Visual Studio Code. Along the way, we’ll discuss what this evolution means for governance, quality, and the role of makers and developers.
Fifteen minutes to pick up a sandwich from nearby shops. We do not have sponsors, so food will not be provided onsite.
Kayode Ajayi
Annabel Okuomose
AI has made it possible for almost anyone to build software. You can describe what you want, ask an AI coding agent to build it, make changes through prompts, and have a working application surprisingly quickly. You no longer need to understand every line of code before you can start creating. But as the application grows, a different set of problems starts to appear. What changed? What did the AI modify? How do you undo something that broke? How do you experiment without affecting the working version? What happens when two people or agents change the same application? And how do you know which version is actually the right one? This is where GitHub becomes useful. This session is a practical introduction to GitHub for vibe coders, AI-assisted developers, low-code developers, and anyone using AI tools to build applications without a traditional software development background. Using a simple application built with AI, we will explore the GitHub concepts that every vibe coder should understand, including repositories, commits, branches and pull requests. Rather than teaching GitHub as a collection of developer commands, the session focuses on the problems these concepts solve. You will see how branches can provide a safe place for you or an AI agent to experiment, how commits create checkpoints you can return to, how pull requests help you review AI-generated changes before accepting them, and how GitHub gives you a history of how your application has evolved. We will also look at how these practices apply across different ways of building software, from AI coding tools and coding agents to low-code platforms such as Microsoft Power Platform. Attendees will see a practical workflow where an application is changed using AI, managed through GitHub, reviewed and safely brought back into the main version of the application. By the end of the session, attendees will understand the GitHub concepts that matter most when building with AI and have a simple workflow they can start using immediately to experiment, manage changes and build with AI more confidently.
Five minutes.
Shubhangi Goyal
What happens when your AI agent encounters unexpected, adversarial, or incomplete inputs? In this session, we explore how to proactively “attack” your own AI systems to make them more robust. We will cover practical red teaming techniques including adversarial prompting, context manipulation, and tool exploitation. Using a hands-on demo in Microsoft Foundry, we will simulate edge cases and stress test an AI agent. This session is designed for practitioners looking to move from experimentation to production-grade AI systems with built-in resilience. Key Takeaways - Simulate real-world adversarial scenarios and edge cases - Stress test agents across context, tools, and decision flows - Identify risks in tool usage, and memory handling - Apply practical guardrails and monitoring using Microsoft Foundry
Five minutes.
Elliott Leighton-Woodruff
Deploying AI agents into production shouldn't be a dark art. Learn how to treat your agent infrastructure with the same care as your application code using Infrastructure as Code to build secure, auditable and version-controlled AI services. In this session, we'll explore the journey from AI Foundry models through API Management to production-deployed agents. You'll discover how adopting IaC practices eliminates ad-hoc deployments and enables seamless environment parity across dev/staging/production, giving your team confidence that every agent deployment is repeatable and reversible. Level: Intermediate to Advanced
Closing notes and a chance to continue the conversation.