Raymond James unveils proprietary AI operations agent

Raymond James is expanding its artificial intelligence push with the launch of a new proprietary digital operations agent designed to give advisors and staff faster answers to day-to-day questions while keeping humans in the loop.

The independent broker-dealer giant said its in-house agent Rai uses natural language processing and generative AI to surface “immediate, curated” guidance from multiple internal knowledge bases, learning over time based on each user’s role, responsibilities and entitlements.

Following a pilot program, the tool is slated to reach select business units first, with an enterprise-wide rollout planned over the coming quarters.

“Rai reflects our strategy of applying artificial intelligence to enhanced service models and secure, scalable applications that empower professionals and financial advisors across the firm,” CEO Paul Shoukry said in a statement Tuesday. “We will continue to deploy our multi-year investment in AI tools and capabilities while steadfastly preserving the personal relationships that remain at the core of our business.”

In practice, Rai functions as an internal chat interface that pulls from Raymond James’ systems and policy documents to support operational decisions. The firm emphasized that the agent is meant to augment staff, not replace them, with “full human-in-the-loop oversight” over how its recommendations are used.

The launch underscores how quickly “agentic AI” is moving from experimentation to production inside broker-dealers. In a blog post yesterday, FINRA flagged a growing set of autonomous or semi-autonomous tools that can interact with staff or clients, write and test code, or orchestrate complex business workflows. While many firms still confine generative AI to basic internal tasks, the regulator highlighted more ambitious agents that are beginning to handle higher-stakes work, from fraud detection to trading-related decisions.

That shift is sharpening questions for compliance and supervision teams. In the post, FINRA executive vice president and head of research and innovation Greg Ruppert highlighted potential risks associated with AI agents shared in FINRA's recent regulatory oversight report, including agents acting beyond a user’s intended authority, opaque multi-step decision paths, and inadvertent exposure of sensitive data. 

Raymond James is positioning Rai within a broader suite of AI tools that already see significant internal adoption. According to Chief AI Officer Stuart Feld, more than 10,000 associates at the firm regularly use conversational AI, and roughly 3.2 million lines of code are generated each month by AI systems under developer oversight.

Amid a broader strategy to invest $1.1 billion annually in technology, the firm said its most recent updates include a CRM note assistant, Zoom meeting summaries integrated with CRM, an internal generative AI search function, and secure access to ChatGPT Enterprise and Microsoft Copilot.

“As an intelligent agent, Rai’s impact will only grow to deliver a bespoke experience for each individual user,” Feld said, adding that the current rollout is “the first of several phases” aimed at integrating Rai across Raymond James’ systems and optimizing processes firmwide.

FINRA, for its part, is inviting firms to share their agentic AI implementations and “proactively engage” with the regulator as strategies develop, framing the discussion as a way to support what Ruppert called the “responsible deployment of these technologies.”