How Nitorum Capital Unlocks Enterprise AI with MCP and Horizon

Meet Nitorum Capital

Nitorum Capital is a forward-thinking investment firm that recognized early on that AI's true value lies not in generic knowledge, but in connecting models to proprietary business context. James Brink, who leads their AI initiatives, saw the gap between consumer AI tools and what businesses actually need—and found Horizon to bridge it.

The Problem with Encyclopedia AI

James Brink describes the evolution of AI in business bluntly: "If we rewind to ChatGPT in 2022, it was sort of like an encyclopedia on the shelf. If you ask a question and it's in the model, great. If it's something that happened after training, it's not going to be there."

Web search helped, but only marginally. "A year or two later, you introduce web search and it's like, well, it's in the encyclopedia or it can quickly find it on the internet. But still, not terribly useful in a business setting because I'm really trying to use the data and things I have in my own business to run my business—not stuff you can Google."

The real unlock, James Brink realized, wasn't about better models. It was about context.

Context Layers: The Missing Piece

"There's so much more that can be accomplished by way of successful context layers within your organization," James Brink explained. "Using MCP to expose those context layers to an agent—that's where the value is."

MCP (Model Context Protocol) lets AI agents access business-specific tools and data. But deploying MCP servers meant learning infrastructure, managing auth, and handling the complexity of server deployment—until Horizon.

Deploying MCP Without the Infrastructure Overhead

Nitorum Capital was an early adopter of Horizon. "We were able to start using the Horizon product early and seeing that we could accomplish what we were trying to do without needing to learn some of the infrastructure of setting up these servers."

The simplicity was striking: "Out of the box, take my GitHub repo of my server, launch it, and use the Horizon OAuth out of the box. It just works for us."

Democratizing AI Development

One of the most powerful aspects of Horizon, according to James Brink, is how it opens up AI development beyond the engineering team.

"If someone in the marketing department or someone in your sales department—someone in whatever department—knows a little bit of Python and can write a workflow, they have the ability to contribute to this. It's not exclusive to engineering."

This democratization has broad implications. "For any business, it's going to be incredibly broad what it can touch. MCP is really core, and Horizon is the avenue by which we can best deploy MCP within our organization."

Key Takeaways

Context beats content: Enterprise AI value comes from connecting models to proprietary business data, not generic web knowledge
MCP is foundational: The Model Context Protocol enables AI agents to access business-specific tools and context
Infrastructure should disappear: Horizon removes the complexity of server setup, auth, and deployment
Python skills unlock participation: Teams beyond engineering can contribute to AI workflows

Lessons for AI Teams

James Brink's advice reflects a broader truth about AI adoption: the technology is only as useful as the context you give it. While consumer AI tools struggle to provide business value, MCP-enabled agents with access to internal systems can transform how organizations operate.

The key is making deployment simple enough that the barrier isn't infrastructure—it's imagination.

Try Horizon Today

Ready to expose your business context to AI agents? Get started with Horizon and deploy your first MCP server in minutes.