Matt Ragone describes how Range financial planners moved from using AI to rewrite messages toward preparing client analyses with agent workflows. Connected client data and standardized advice procedures give those workflows context, while a financial planner reviews the output before it reaches a client.
Matt Ragone outlines five steps: pick a repeatable workflow, teach the agent with real cases, package the instructions as a skillAn AI agent skill is a reusable package of instructions, resources, and tool guidance for performing a bounded kind of work., test standard and edge casesEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria., and ship it with an owner and reviewer. He recommends building in questions for missing context and explicit limits on sending adviceAn agent permission boundary defines and enforces which resources and actions an automated agent is allowed to use..
Matt Ragone says the company shares skills across teams and that subject-matter experts should help build and maintain them. He argues that the next challenge is keeping the growing collection tested and current.
Watch on YouTube
