Firms can now build governed AI agents in Zeplyn and access them through Claude, ChatGPT, Copilot, Slack, and proprietary applications

Zeplyn, the AI operating system for wealth management, has released Zeplyn Agent Studio and the Zeplyn Model Context Protocol (MCP) Server, two new capabilities that enable wealth management firms to create custom AI agents in Zeplyn and make them available through the AI applications their teams already use. 

Agent Studio puts firms in control of how AI applies their expertise. Using plain-language instructions, wealth management firms can define how an agent should operate, what information it should consider, and what its output should look like. Firms could create an agent that prepares annual review briefs according to its standards, produces meeting recaps in a consistent structure, or drafts client emails in the firm’s voice. Once published, the agent becomes available across the firm, helping the firm create greater consistency while preserving its own approach to serving clients. 

“When I’m having a touchpoint call with a client, I want to focus on the relationship, not spend time digging through notes for the one personal detail I know they shared with me,” said Monica Cervantes, financial advisor at PAX Financial Group. “Clients tell us so much over the years. With Agent Studio, we created an agent that pulls those non-financial details into one view, including direct quotes from the client. It helps me walk into conversations prepared to pick up where we left off and continue building the warmth and trust we value in our client relationships.” 

The Zeplyn MCP Server extends those agents beyond the Zeplyn interface through a standard protocol. It gives MCP-compatible AI applications and proprietary advisor tools governed access to agents created in Agent Studio, Zeplyn’s client intelligence, and more than 60 wealth-specific tools. This allows firms to apply the same intelligence across different parts of their technology environment without building and maintaining a separate connection for each application. 

For example, an advisor working in Claude could request a preparation brief for a client household. Through the Zeplyn MCP Server, Claude could securely retrieve recent meeting summaries, CRM information, financial planning data, and current holdings from Zeplyn and then compile them into a single brief. 

“Every wealth firm has expertise that shapes how it serves clients, and AI should help put that expertise to work at scale,” said Era Jain, co-founder and CEO of Zeplyn. “With Agent Studio and MCP, firms gain AI infrastructure they can build on, allowing them to shape that intelligence as their business evolves and bring it into the tools their people choose to use.” 

Both capabilities are available now and operate within Zeplyn’s existing governance architecture, which includes role-based access, audit trails, PII redaction, configurable data retention, and human review. Client data is not used to train underlying models. 

To learn more about Zeplyn, visit zeplyn.ai.