1. What MCP means in practical terms
MCP was presented as a delivery method built for AI systems like Claude, ChatGPT, Gemini, and Perplexity. The key idea is that AI does not simply want raw endpoints or raw datasets — it works best when content is exposed as clearly described “tools” in structured JSON, so it can understand what is available, choose the right tools, and reason over the results.
Rather than thinking of MCP as “just another API,” the meeting framed it as a standardized way to make TC’s content usable inside broker AI assistants, copilots, conversational screening tools, and even AI-generated webpages or dashboards.
2. Why this matters to our AI strategy
The core strategic message was that TC wants to be the one-stop shop for investment intelligence in client AI systems. The MCP already exposes a broad set of TC capabilities across multiple verticals, including technicals, fundamentals, options, price analysis, Market Buzz, economic content, and Storyteller. The transcript references 51 tools across 8 domains, all available through one connection.
That breadth matters because brokers do not want to stitch together many disconnected MCPs if they can avoid it. The meeting positioned TC as the connector that can reduce fragmentation and help clients avoid the risks and complexity of mixing too many separate AI data sources.
3. Why TC adds value beyond raw data
A major point in the discussion was that clients need more than raw market data if they want to create useful AI experiences for investors. TC brings methodology, consistency, and advanced analysis layers — not just facts. That includes pattern recognition, sentiment analytics, technical signals, quantamental scoring, and narrative explanation.
The meeting also highlighted that this approach can make AI outputs cheaper, faster, and more consistent, because TC provides a dense intelligence layer that reduces the amount of raw information an LLM must process. That helps with AI context-window limits and supports more reliable answers than asking the AI to independently analyze everything from scratch.
4. The trust, compliance, and quality angle
One of the most compelling points for clients is that MCP allows brokers to ground their AI experiences in licensed, high-quality, analyst-backed content rather than relying on random web sources. The speakers explicitly connected MCP to reduced hallucination risk and better explainability, since brokers can point to a known, trusted provider with a long-established methodology.
This is especially important when compliance teams ask where the data and analysis are coming from. The message was clear: TC gives clients a credible, established intelligence source for their AI workflows.
5. Where Storyteller fits
The meeting reinforced that Storyteller complements MCP in two ways. First, it gives clients a faster on-ramp when they are not yet building full AI projects. They can start using AI-generated summaries and narratives now through more traditional UI experiences. Second, Storyteller content is also available as part of the MCP dataset, so it can feed broker AI systems too.
This creates a strong paired message for clients:
6. What clients can do with it
The examples shared showed that MCP can support much more than chat. Brokers can use it for conversational screening, AI-generated dashboards, personalized research pages, and workflows that combine TC intelligence with their own customer data, holdings, and suitability context.
The discussion also introduced the idea of the broker controlling the orchestration layer: they decide which tools are used, how they are prioritized, what other proprietary data is mixed in, and what compliance or tone guidance wraps the final response. That is an important distinction: TC provides the intelligence layer and AI-ready tool structure, while the broker shapes the end-user experience.
7. Practical sales takeaway
The commercial takeaway was that we should show clients the whole toolkit, not just one narrow use case. Even if they are not yet using TC content in their usual widget or iframe formats, they may still benefit from TC as a foundational AI dataset. The transcript makes clear that the goal is to position TC as the source powering these projects.
The pricing discussion also suggested a packaging approach more similar to other products: a monthly fee with an included number of MCP calls, plus overage usage fees. The intent is to be more predictable and less intimidating than a purely variable-cost model.
8. Suggested conversation starters for clients
A few strong follow-up angles come out of the meeting: