20 minutes. 20 seconds.
That's how long it used to take a salesperson to research a target list of allocators, versus how long it takes now using MCP. The difference isn't faster typing or better dashboards. It's an entirely new way for live data to move between systems.
MCP is newer than most of the AI tools fundraisers are already using, and it changes who in your organization can actually pull data from sources like Dakota Marketplace. The short version: anyone with access to an AI assistant, with no developer required.
This article covers what MCP is, what the Dakota MCP server does, and how it works alongside the Dakota API.
What MCP Actually Is
MCP stands for Model Context Protocol. It's an open standard created by Anthropic, the company behind Claude, that lets an AI assistant connect to live data sources in real time.
The way to picture it: an MCP server sits on top of an existing data source (usually accessed via an API) and translates between two languages. On one side, the AI assistant is asking questions in plain English. On the other, the data source is responding in structured records. The MCP server is the translator that makes those two sides talk.
For a fundraiser or allocator, the practical result is this: you can connect Claude, ChatGPT, Gemini, or any other compatible AI assistant to the Dakota MCP server, then ask questions of Dakota Marketplace data as if you were typing them into a chat window. The AI assistant pulls the data live, formats it, and gives you back exactly what you asked for.
No code. No custom queries. No exporting CSVs and re-uploading.
What the Dakota MCP Server Actually Does
The Dakota MCP server is live today. Connecting it to your preferred AI assistant takes under two minutes, with no developer involvement.
Once connected, your team can ask questions like:
- "Show me all endowments in the Northeast that invest in private equity."
- "Pull a list of family offices in Texas that allocate to private credit."
- "Which institutional investors made recent commitments to infrastructure?"
- "What's the most recent fund commitment activity from Texas Teachers' Retirement?"
Each query returns live Dakota data in seconds. From there, your team can keep working inside the same AI tool: drafting outreach, creating follow-up tasks, feeding the results into other automations. You never leave the tab. The experience parallels Dakota Joe, Dakota's natural-language tool for CRM reporting.
The Dakota MCP server uses the same security model as the rest of Dakota Marketplace. Whatever you have access to in Marketplace is exactly what you can pull through the MCP server. If you subscribe to Dakota FAs but not the full Marketplace, your AI assistant returns the FA data you're licensed for, with no spillover into other datasets. Authentication happens inside your own tenant, and your prompts are not fed back to Dakota.
See the Dakota MCP server in action. Connect it to Claude, ChatGPT, Gemini, or Copilot in under two minutes and start asking questions of live Dakota data. Book a demo.
How MCP and the API Work Together
MCP and the Dakota API are not competing technologies. The MCP server sits on top of an API and uses the same underlying data.
The choice between them comes down to who's doing the work and what they're building. The Dakota API is the right path when your engineering team is building a custom integration, an internal dashboard, or a proprietary tool that needs programmatic access to a specific slice of Dakota's dataset. The MCP server is the right path when fundraisers, IR teams, or anyone else without code skills wants to query live data directly from an AI assistant.
Most fundraising teams will use both. The API gives your engineering team programmatic access to push live Dakota data into your CRM, data warehouse, or analytics platforms. The MCP server gives anyone the ability to ask Dakota a question in plain English and get a live answer back.
See Dakota's MCP server in your AI assistant
Dakota Marketplace covers accounts and contacts globally, updated daily, with 75+ investment preference filters across every allocator type. Connect the Dakota MCP server to Claude, ChatGPT, Gemini, or Copilot in under two minutes and ask live questions of the same data your CRM relies on.
Book a demo.
