Genesis Global, the application development framework provider, has introduced a Model Context Protocol (MCP) Server to provide a controlled and compliant interface between large language models (LLMs) and enterprise applications built on its platform.
The MCP Server supports the emerging open MCP standard, which aims to unify how software applications deliver operational and contextual data to LLMs. Its growing adoption in the financial sector reflects demand for more practical and secure AI integration, particularly in data-sensitive environments.
Designed as an optional component for Genesis platform versions 8.11 and above, the MCP Server enables selective AI-driven actions within applications without requiring any reconfiguration. It functions as a gateway, allowing approved AI tools to discover and interact with Genesis-built applications under strict governance rules.
“The agentic aspect of MCP is what’s really gaining traction,” says Tej Sidhu, CTO of Genesis Global, in conversation with TradingTech Insight. “Just as REST transformed how we consume APIs, MCP allows systems to describe the services they offer and the metadata around them, making it possible for an LLM to understand and orchestrate application functionality. From version 8.11 onward, any application built on Genesis can automatically expose an MCP interface. What’s particularly exciting is how some forward-looking banks, starting with their operations teams, are aiming to shift the primary user interface away from traditional applications and toward AI agents such as internal chatbots. With MCP, those agents can now interact directly with applications, execute tasks, and even coordinate multi-step business processes, all from within a conversational interface. That’s a compelling shift in how people engage with enterprise software.”
Application owners retain full control over which functions are exposed to AI agents and can enforce existing entitlements and permissions. Human-in-the-loop features allow for oversight and approval of AI-initiated actions where necessary.
“We’ve built MCP on top of Genesis’ robust entitlements framework,” notes Andrew Rendell, VP of Platform Engineering at Genesis. “Our platform already supports complex user hierarchies and strict compliance requirements, ensuring, for example, that one user group can’t access another’s data. That same governance carries over to MCP. Every piece of data has entitlements attached, specifying exactly who can see it. This level of control is often overlooked when systems are built quickly, but it’s critical for secure AI integration. It’s not just about single sign-on or high-level access, it’s about embedding fine-grained visibility and compliance directly into the data layer.”
Initial use cases for the MCP Server have included data extraction for LLM processing, conversational interfaces for user interaction, and orchestration of multi-application workflows using AI.
The MCP Server forms part of Genesis’s broader effort to ensure AI is a predictable and compliant component within enterprise technology stacks.
“Looking ahead, the MCP Server is going to be absolutely critical,” predicts Sidhu. “Banks have invested heavily in becoming API-first, but building standardised APIs across all systems remains a huge lift. MCP plays a similar role to what REST did for APIs, especially as firms now expect every part of their stack to be AI-aware. One of the big challenges with LLMs is that they don’t inherently understand a firm’s internal data. You can spend significant time and resources trying to train them, but MCP offers a simpler path. It’s a straightforward, open standard, and the reason it’s gaining traction is because the market clearly needs it. In just a month or two, we’ve gone from MCP being introduced to seeing nearly every senior IT leader running AI initiatives talking about it. That tells me it’s going to be foundational.”
Subscribe to our newsletter
