Nicole Clarke, the CEO and a co-founder of Trellis shared on LinkedIn that they’ve been overwhelmed by their customers’ excitement by being able to access their Trellis data within Claude. 

One customer took their excitement about the Claude integration a step higher.

Reading what I have the last two days about an MCP connection to Claude makes me think companies serving up data are going find Claude a gateway to customers and their satisfaction.

When I raised an MCP for The Library by LexBlog, I learned the team was already ahead of me. They were already exploring an MCP connector and had begun the API work that could help support it.

Basically, from the tech company’s standpoint, deploying an MCP server involves exposing ones AI tools, resources, and prompts over a standardized connection (usually Streamable HTTP or SSE) so that models like Claude or Gemini can access them securely.

During the advent of the web, I shared it was like being on a surfboard and trying to stay on top of the wave. AI feels like a tsunmai.