You may have seen the recent hype surrounding MCP and MCP servers, but if you haven’t, allow us to explain.
What is an MCP Server?
Model Context Protocol is a new standardised way for an artificial intelligence (AI) system to integrate with external tools, systems and data sources.
This opens powerful additional capabilities for an artificial intelligence user. As you can now conveniently get data from your supported system, your AI agent can process it however you desire. Additionally, you can run any supported functionality too.
An example of an MCP Server.
While reviewing the list of available MCP servers, we noticed Xero. Given our recent partnership (read about it here), we were eager to experiment, recognising the huge potential of processing Xero data through an AI agent. We set up a demo company, configured the developer backend, connected the AI agent, and the integration was ready to go.
We asked our AI agent ‘Can you give me sales projection charts that I can give to my sales department that show how much we need to sell to make an extra 15% profit within the next 12 months’
Why is MCP coined as the USB-C port for all AI applications?
You may have heard MCP described as the “USB-C port” for AI agents. The analogy works because, just as USB-C has become a universal, standardised connection for hardware devices, MCP provides a universal, standardised way for AI agents to connect and interact with external tools and systems. In the same way you can plug almost any modern device into a USB-C port and expect it to work, MCP allows AI agents to integrate with a wide range of capabilities consistently and predictably.
How can we utilise MCP to our advantage?
The MCP server has many similarities to a traditional REST API. Therefore, through an experimental process, we created an MCP server for our client Scrub Club (read about it here). Now, by utilising their AI agent, the admins of Scrub Club can not only analyse the data from external tools but also data from their internal database.
Additionally, by incorporating an LLM (large language model) into your software and linking your external systems through MCP, you open a wide range of features and possibilities to enhance your application. Scrub Club intend to take advantage of this by enhancing its use of an existing OpenAI LLM and running commands based on its customers’ input, furthering its brilliant use of automation.
Case Study: How a custom MCP Server unlocks the power of data analysis
Read about how a new custom MCP Server can help your business stay current with the latest AI advancements and unlock powerful data analysis
Can any business set up an MCP Server?
Yes. A custom-built MCP server lets any business securely open up their systems—whether to extract data, trigger processes, or integrate with AI-driven workflows.
We develop bespoke MCP servers tailored to your systems and security needs.
How secure is an MCP Server?
Security depends on the implementation.
We design MCP servers with enterprise-grade protection, following best practices in authentication, encryption, and access control—ensuring your data stays safe and your systems remain secure from attacks.