Hackathon Success Story: How One OpenEdge Developer Built an AI Copilot in Two Weeks

The OpenEdge × AI Innovation Challenge invited developers, partners and independent teams to explore how artificial intelligence could solve real business problems using Progress OpenEdge as the trusted foundation for enterprise data, logic and operations.

For Michael Whelan, the hackathon became more than a short-term challenge. It became the catalyst for a much larger transformation in how he thought about OpenEdge, AI and the future of enterprise application development. When the hackathon began, Michael Whelan described himself as new to modern AI tools and Visual Studio Code, but he used the experience to quickly learn, experiment and build. “Within a few months, I went from a dinosaur to being a spaceman,” he said.

What started as a learning opportunity soon evolved into a practical AI-enabled solution that connects enterprise systems, exposes existing business logic and helps users get answers from complex data environments more securely and efficiently.

Seeing AI as a New Way to Use Existing Business Logic

Like many experienced OpenEdge developers, Michael Whelan was working in an environment with long-standing enterprise systems, custom code and business processes that still played a critical role in day-to-day operations. He explained that many developers were using older tools and workflows to query databases, compile code and investigate business issues, while custom code lacked some of the modern review and repository practices organizations rely on today.

During the hackathon, Michael Whelan challenged himself to learn new tools and see what was possible with AI. At first, the experience was less about building a finished product and more about learning how to use the tools. But that learning quickly paid off. He began using AI assistance in his daily work and saw firsthand how quickly these tools could help him modernize his development approach.

The real opportunity became clear when business users continued asking practical operational questions. Could IT check a purchase order? Why had something not loaded? Had an order shipped, been rejected or been delayed? Instead of manually investigating these requests across multiple systems, Michael Whelan saw an opportunity to use AI as a new interface into trusted enterprise logic and data.

Connecting AI to Enterprise Systems Through OpenEdge

Michael Whelan built his solution on top of the OpenEdge MCP Server—a layer that allows AI agents to connect securely to existing PAS for OpenEdge services (the application server that runs OpenEdge business logic) without providing the AI direct access to the underlying database. Those services can then securely access the enterprise systems and business processes already in place. In his environment, that meant connecting across multiple ERP and data systems, including QAD (an ERP platform built on OpenEdge) and other enterprise platforms.

Rather than giving AI direct access to databases, the solution keeps control in the application layer. Michael Whelan explained that the OpenEdge MCP layer handles authentication, security and controlled access to tools and services. “The golden box is the MCP,” he said, because without it there is no secure way to connect AI to the business logic and tools exposed through PAS for OpenEdge.

This approach is important because it allows AI to work with enterprise systems without bypassing the safeguards those systems require. Michael Whelan emphasized that data security, governance and access control are essential. His implementation was designed so data remains inside the network, transport is encrypted and interactions are structured as JSON requests and responses rather than unrestricted database queries.

From Purchase Order Questions to Real Business Workflows

The solution began with a practical business use case: helping users understand what happened to a purchase order. A user could ask a question in natural language, and the AI-enabled experience could determine whether the order had shipped, whether it had been invoiced, what shipment number applied and whether related documents were available.

From there, the workflow continued to expand. Michael Whelan described how users could ask to see order lines, retrieve invoice details or request that a copy of an invoice be emailed. The request would then be submitted through PAS for OpenEdge, with existing ERP processes handling the action behind the scenes.

That connection between AI and existing systems is what makes the project powerful. The AI experience is not replacing the enterprise application. It is making the existing application easier to interact with. By exposing business logic as tools through OpenEdge and MCP, Michael Whelan created a way for users to ask business questions in natural language while keeping the underlying processes governed and secure.

The result is a practical path for organizations that rely on long-standing enterprise applications. Instead of replacing systems that still work, organizations can modernize how users interact with them. As Michael Whelan explained, companies often spend significant time and money replacing systems when they may be able to secure them, expose the right services and provide a more modern interface on top.

Building Quickly by Reusing What Already Exists

One of the most compelling parts of Michael Whelan’s story is how quickly the solution came together. Once the core tools were in place, he said the project took about two weeks at most, and possibly closer to one week, because he already had PAS for OpenEdge connectors to the ERP systems. The work was primarily about adding the tool layer that allowed AI to ask the right questions and retrieve the right information.

That distinction matters. The project did not require rebuilding the enterprise environment from scratch. Instead, it reused existing services and business logic, then exposed them in a new way through MCP and AI.

Michael Whelan designed the experience so the AI does not retrieve everything at once and decide what matters. Instead, he separated common user questions into focused tools that could return specific responses quickly. If more information is needed, the user can ask for more.

This reflects an important lesson for enterprise AI: success depends not only on connecting to data, but on structuring access in a way that is purposeful, secure and aligned to business workflows.

Making AI Practical, Secure and Useful

For Michael Whelan, the project reinforced that AI should not be treated as a shortcut around enterprise architecture. It should be connected thoughtfully to the systems, rules and controls that already govern the business.

He was especially focused on making sure AI did not expose sensitive information or create unnecessary risk. He noted that simply having a connector does not mean it should be broadly exposed, and developers need to think carefully about who has access and where data is going.

That security-first mindset is one reason the solution has attracted attention inside his organization. Michael Whelan described sharing the project with IT leadership and said the value was clear because the approach recognized the importance of security, data governance and keeping information protected.

At the same time, the project shows how AI can deliver meaningful business value. It can help customer service teams answer questions faster, make it easier to understand order and invoice status, and create a more natural way for users to interact with complex systems.

Advice for Developers Getting Started with AI

When asked what advice he would give to other OpenEdge developers who are curious about AI but unsure where to begin, Michael Whelan pointed back to the hackathon experience. Events like the OpenEdge × AI Innovation Challenge can provide motivation, structure and access to others who are learning or building.

He also encouraged developers not to try to learn everything alone. Joining a group, asking for help and learning from others can make the process easier. He recommended using available resources, training and community channels, including Discord, to find guidance, development containers and practical support.

Most importantly, Michael Whelan encouraged developers not to be intimidated by AI. In his view, AI is another interface into business logic, much like character interfaces, GUIs or .NET front ends before it. The front end may change, but the business logic and middleware remain essential.

That perspective is especially relevant for OpenEdge developers. The skills they have built over years of designing business applications, structuring middleware and exposing reliable services are still highly valuable in an AI-enabled world.

Why It Matters

Michael Whelan’s story demonstrates that AI innovation does not always require replacing core business systems or starting from scratch. For organizations that depend on OpenEdge applications, AI can become a new way to unlock value from the systems, services and data they already trust.

By using OpenEdge, PAS for OpenEdge and the OpenEdge MCP Server, developers can expose business logic to AI in a controlled, secure and practical way. That creates opportunities to modernize user experiences, streamline operational workflows and help employees get answers faster without moving critical enterprise knowledge outside established systems.

The OpenEdge × AI Innovation Challenge helped Michael Whelan see what was possible. But his work after the hackathon shows something even more important: with the right architecture, AI can become part of real enterprise workflows today.

As he put it, the skills OpenEdge developers have learned over time still matter. “The skills I learned then still serve me well today, and the skills I learned today are going to serve me well tomorrow.”

Discover What’s Possible with Progress OpenEdge

The OpenEdge × AI Innovation Challenge showcased how developers are using OpenEdge alongside modern AI technologies to solve real business problems. Learn more about Progress OpenEdge, the OpenEdge AI Assistant and the OpenEdge MCP Server to discover how you can bring AI capabilities to your own OpenEdge applications.