This Level Up introduced Trading Central’s MCP server as a new, AI-ready way for clients to access Trading Central research and analytics. The immediate goal is to enable field representatives to demonstrate MCP in Claude or ChatGPT, uncover and qualify customer AI projects, and identify where Trading Central can provide trusted, consistent intelligence.
Top 10 takeaways
- The goal is to uncover and qualify real customer projects.
Use the demo to learn what the client is building, who owns the initiative, which users it serves, what intelligence it needs and whether there is a path to a proof of concept. - Demo the value of multiple data sets from one connection.
Connect Trading Central MCP to Claude or ChatGPT and demonstrate a relevant asset briefing, comparison, or visual dashboard. Instruct the assistant to use Trading Central tools and highlight multiple tools from one access point. - Trading Central’s differentiation is trusted, consistent and actionable intelligence.
Our licensed and traceable research, established methodologies and consistency across client experiences help reduce contradictory or unsupported AI responses. It also reflects decades of experience delivering independent, third-party research that turns complex market information into actionable insights that support investor decisions. - Know when to use an API versus MCP.
Use an API when developers want to select specific data, apply fixed business logic and control exactly how it is presented. Use MCP when an AI needs to discover, select and combine Trading Central tools dynamically based on the user’s request. - MCP and APIs expose the same underlying Trading Central intelligence.
They are different delivery methods rather than different data sets. Clients generally should not recreate Trading Central’s MCP over its APIs because Trading Central maintains and continuously improves the tools and their descriptions. - Understand Storyteller versus MCP.
Storyteller is a Trading Central product that generates a consistent, ready-to-display narrative. MCP is an access layer that allows the client’s AI to retrieve Storyteller content alongside other Trading Central tools and decide how to use it. Storyteller can therefore be delivered through an API or accessed as a tool through MCP. - Claude and ChatGPT are demonstration and exploration environments.
They help reps and clients interact directly with the data and envision possibilities. A broker’s production solution may instead use its own LLM, agent framework, interface and governance controls. - Set expectations about access, costs and scope.
Clients receive tools according to their entitlements though our focus is on pitching one stop shop for all tools. Usage is measured similarly to API usage, although one AI request may generate several tool calls. Usage caps can help clients manage experimentation costs. - Use the Employee Hub as the source for current enablement resources.
Before demonstrating MCP, field reps should review the battle card, client demo playbook, developer integration guide, and other current materials in the Employee Hub. They should also spend time testing the recommended prompts themselves. - Know when to involve technical specialists.
Field reps should lead initial discovery and demonstrations. Once the client has a defined use case and wants to discuss use cases, architecture, tool selection, permissions, output controls, production deployment or detailed usage, bring in Fayez, Jerome, Kathryn, or the CS Tech team as needed.
