From Geeky Idea to Enterprise App: Comparing App Builder, Power Apps Vibe, and Code Apps

I have a geeky hobby: dynasty fantasy football. It’s a long‑running game where you build virtual teams made up of real NFL players, track performance over multiple seasons, and constantly make decisions based on partial data, rankings, and gut feeling.

Over time, this evolved into a very familiar problem, one that shows up in many business scenarios as well:
I needed one place to capture decisions, compare them against external benchmarks, and learn from historical outcomes across years.

This session takes that exact app concept and builds it three different ways, using three Microsoft app‑building approaches—each optimized for a different set of tradeoffs:

Approach 1 — App Builder in Microsoft 365 Copilot
A fast, conversational way to build lightweight apps backed by Microsoft Lists. Ideal for quick insights, personal productivity, and simple sharing, but with clear boundaries once structure and scale increase.

Approach 2 — Power Apps Vibe
An AI‑native Power Apps experience that generates the solution holistically: requirements, Dataverse data model, and a working app with generated code. Designed for more structured scenarios where scalability, relationships, and extensibility matter.

Approach 3 — Code Apps
The pro‑developer route, offering full control over UI, architecture, ALM, and extensibility on a governed platform. This approach won’t be demoed live, but will be explained architecturally to highlight where and why it becomes the right choice.

The session is demo‑heavy for App Builder and Power Apps Vibe, and architecture‑ and decision‑framework‑focused for Code Apps. Along the way, we’ll explore:
- How AI‑generated apps differ depending on the underlying platform
- The technical and governance implications of Lists vs. Dataverse
- How to explain these tradeoffs to stakeholders who “just want an app by tomorrow”

You’ll leave with a practical mental model for why these tools exist, when to use each one, and how to choose the right approach, before rebuilding the same app three times for the wrong reasons.

Inside a Production-Ready Sales Agent: Copilot Studio, Dataverse & MCP in Action

AI agents become truly valuable when they move beyond demos and start supporting real work, especially for roles like sales managers who are constantly on the move.

In this demo-heavy, technical session, we’ll guide you through building an AI-powered Sales Agent using Microsoft Copilot Studio and the Power Platform, based on a relatable real-world scenario: a mobile sales manager navigating daily tasks while on the road.

This session is aimed at Power Platform makers, Copilot Studio builders, and technical consultants who already have basic experience with Dataverse and the Power Platform and want to design more advanced, business-driven AI agents.

We’ll start by framing the business case: why AI-driven solutions matter for mobile sales teams and where Copilot Studio fits within a modern sales architecture. From there, we’ll move into a live build, showing how the individual components come together:

- Using a Model-Driven App as the foundation for a scalable sales solution.
- Building a custom Copilot Agent and enriching it with structured business data using Dataverse.
- Understanding Knowledge vs. Model Context Protocol (MCP) - what MCP is, when to use it, and how it enables real-time automation.
- Automating real-world sales tasks, such as promoting quotes to orders and generating documents, helping sales managers reduce repetitive work while staying productive on the go.

By the end of this session, you will be able to:

- Design and deploy a conversational AI agent tailored to specific business scenarios.
- Integrate Copilot Studio with Dataverse, model-driven apps, and other Power Platform components.
- Apply MCP to enable advanced automation and real-time data access.
- Deliver practical AI-driven solutions that optimize sales workflows and improve day-to-day productivity.

Whether you’re building your first Copilot agent or looking to push its capabilities further, this session will give you the technical insights and practical patterns needed to create smarter, more effective business solutions.

AI Agents That Get Work Done: The Operating Model to Scale Trusted Digital Teammates

AI agents aren’t just the future, they’re a real advantage when they move beyond answering questions and start getting work done. The hard part isn’t the tech. It’s how you make agentic solutions succeed inside a real business: the operating model, the trust model, and the enablement steps that turn pilots into outcomes.

In this session, I’ll share the framework and operating model I use to turn agentic solutions into successful digital teammates, grounded in practical lessons from the field.
This session is for product owners, architects, makers, and adoption leads who already understand what agents are and want to scale them responsibly across teams.

You’ll learn:

- How to evolve from simple Q&A bots to task-based agents that reliably deliver measurable value.
- Why human-in-the-loop is your secret weapon for trust, adoption, and quality, especially when agents touch real processes.
- Which business enablement steps make or break rollout, including ownership, decision points, escalation paths, and the habits teams need to actually use these AI teammates.

You’ll also see concrete examples that go beyond the typical “Q&A HR agent,” including agentic Request for Proposal (RFP) solutions that streamline end-to-end processes and Request for Quote (RFQ) automation that still keeps humans in control at the right moments. Expect a mix of case studies, pitfalls to avoid, and a practical framework you can take home and apply to your own environment.

We will not dive deep into foundational “what is an agent” theory or do step-by-step tool setup. This session focuses on the operating model and rollout mechanics that determine whether agents succeed in real teams.

If you’re ready to unlock the real potential of AI agents, and make them work for your teams, this session is for you.

Trust by Design: Building Human-in-the-Loop AI Agents with Copilot Studio & Power Platform

Keeping humans in the loop isn’t just a best practice, it’s the difference between trusted AI adoption and operational chaos. As AI agents become more autonomous, designing clear control points for people becomes essential.

In this demo-heavy session, you’ll see how to design human-in-the-loop agentic solutions using Microsoft Copilot Studio and the Power Platform, with practical patterns you can apply immediately in real-world environments.

This session is aimed at Power Platform makers, Copilot Studio builders, architects, and technical decision-makers who are already familiar with building copilots or cloud flows and want to move beyond simple automation toward responsible, production-ready AI agents.

You will learn how to:

- Monitor and supervise agent behavior using the Agent Feed to maintain real-time visibility into conversations and decisions.
- Design approval and intervention patterns with Advanced Approvals and Request for Information flows, bringing humans back into critical decision points.
- Extend Copilot Studio with custom Power Platform solutions to align agent behavior with your organization’s governance and risk model.

We will focus on practical design patterns, demos, and architectural trade-offs.
We will not cover introductory Copilot Studio concepts, AI theory, or organization-wide adoption strategies.

If you want to build AI agents that are powerful, transparent, and accountable, without slowing teams down, this session gives you a clear, proven blueprint.

When Copilot Scales Faster Than We Do

Launching Copilot worked for our pilot customer groups. The moment we scaled beyond them, our approach started to fail.

Our early pilots succeeded because they involved motivated, self‑selecting users. That success did not survive broader rollout: skill levels diverged, expectations shifted, and “helpful AI” became a new layer of operational reality. When agents entered the picture, individual experimentation turned into production‑level impact, often without anyone clearly owning the outcome.

And here’s the hard part we didn’t expect: we couldn’t keep up either. As consultants, we were delivering guidance in a project rhythm while the technology and behaviors evolved weekly. Our own teams started repeating the same explanations and fixes, and customers got stuck in a frustrating loop without meaningful maturity gains.

We stopped treating adoption as a series of projects and reorganized around a continuous program with shared ownership across teams. We will share what broke at scale, the decisions we reversed, and the practices we no longer use, both with customers and inside our own consultancy. Along the way, we’ll show how we now scale knowledge inside our teams, create ownership to help the program improve (instead of “someone else will fix it”), and put guardrails in place that enable innovation rather than shutting it down.

This is a field report from the messy middle of AI adoption, where scaling people turns out to be harder than scaling technology.

What attendees will take away

- Why pilot success often collapses at scale and the assumptions that cause it
- What changed when agents made “everyone a builder,” and why ownership becomes the real bottleneck
- How we shifted from project delivery to a continuous, maturity‑driven model and what that required internally
- Practical patterns for scaling knowledge across teams without centralizing everything (and without burning out your experts)
- The mistakes we made in training + governance timing and how we corrected them

Level: Intermediate (200–300)
Audience: Consultants, IT Pros, Architects, Adoption & Platform Leads