Nymbus launches MCP server to connect AI with core banking

By Vriti Gothi

Nymbus has introduced a Model Context Protocol (MCP) server designed to connect artificial intelligence (AI) assistants directly to core banking systems, as financial institutions seek more controlled ways to deploy AI in operational environments.

Announced on 9 April, the Nymbus MCP Server is among the first purpose-built MCP implementations for core banking. It provides banks and credit unions with a standardised interface to integrate AI-powered tools into front-office operations while maintaining governance over access, workflows, and data usage.

The server currently supports 19 tools covering common banking functions, including customer verification, account management, payments, and debit card controls. These tools enable AI assistants to execute approved actions via large language models, allowing staff to complete tasks through conversational interfaces rather than navigating multiple systems.

The development reflects a broader industry shift from AI experimentation to embedded, workflow-level deployment. Financial institutions are increasingly looking to apply AI to operational processes such as fraud investigations, case management, and customer servicing, where efficiency gains can be realised without compromising oversight.

“AI creates real value in banking when it helps institutions get work done, not just generate answers,” said Jeffery Kendall, Chairman and CEO of Nymbus. “With the Nymbus MCP Server, we are giving banks and credit unions a practical way to put AI into everyday workflows while maintaining the control, consistency, and accountability they need to operate with confidence. Our server simplifies common front-office work today while laying the groundwork for more intelligent, agentic banking experiences in the future.”

Unlike legacy core systems that often require bespoke integrations for each AI use case, the MCP server introduces a unified connection layer. This architecture is designed to reduce integration complexity, accelerate deployment timelines, and enable institutions to scale AI-assisted workflows more consistently across teams.

The platform incorporates security and compliance features aligned with banking requirements, including token-based authentication, role-based access controls, encrypted connections, audit logging, and masking of personally identifiable information in logs. Institutions retain granular control over which tools are enabled and how workflows are approved and executed.

By building on the open MCP standard, the launch also signals a move towards more interoperable AI ecosystems in banking. As the sector explores agentic AI models capable of executing multi-step tasks, standardised protocols such as MCP could play a key role in enabling secure, scalable integration between AI systems and regulated financial infrastructure.