Data sourced from Dakota Marketplace, the global LP and GP intelligence platform trusted by thousands of investment professionals. Learn More | Book a Demo
For fundraisers, sales teams, and investor relations professionals who spend part of every day inside an AI assistant, Dakota's MCP connector removes the tab switch. The same allocator profiles, contacts, and mandate detail that live inside Dakota Marketplace are now accessible through Claude, ChatGPT, Copilot, with no code required.
This post covers what the connector does, how it changes the fundraiser's day, and which AI tools it works with today.
What the Dakota MCP Connector Is
The connector is a Dakota-built MCP server that sits on top of our existing API. MCP, or Model Context Protocol, is an open standard created by Anthropic that lets AI assistants connect to live data sources in real time. Once your team's AI assistant is connected to the Dakota MCP server, you can ask questions in plain English and get answers pulled straight from Marketplace.
No SQL. No filter menus. No exports. You type the question the way you'd ask a colleague, and the AI assistant returns the answer with live Dakota data behind it.
Setup takes under two minutes on any of the supported AI assistants. Once configured, the connection works across web, desktop, and mobile.
How It Changes the Fundraiser's Day
The most concrete way to describe the shift is by task. Here's what changes when Dakota data is one prompt away instead of one tab away.
- Building target lists. "Pull a list of family offices in Texas that allocate to private credit, with investment-team contacts." The list comes back in seconds. You can refine it in the same chat: add a size filter, exclude firms you've already met with, sort by recent commitment activity.
- Prepping for meetings. Before a call with an allocator, ask your AI assistant for a summary of the firm's recent commitments, current allocation mix, and any recent news. The context that used to require three tabs and a spreadsheet arrives in one message.
- Prepping for trips. City scheduling gets faster when you can ask "who should I see in Miami next Tuesday and Wednesday" and get a list of relevant allocators, ranked by fit and priority, without opening a separate database.
- Ad-hoc research. Market questions that used to sit at the bottom of your to-do list get answered in seconds. "Which pension funds have committed to infrastructure in the last six months?" "How many endowments in the Northeast allocate to hedge funds?" No queue, no formal research request.
- Drafting outreach with live data as context. Once you have the target list in your chat, you can ask your AI assistant to draft personalized intro emails using each prospect's specific allocation profile. The context and the drafting happen in one place, with no copy-paste between windows.
- Cross-checking coverage. Before a conference, ask your AI assistant which allocators on the attendee list your firm has already met with and which are net-new. Turn a two-hour manual audit into a five-minute question.
None of these tasks are new. What changes is how long they take and how many tabs they require. The research still happens; it just stops being a bottleneck.
Which AI Tools Work with the Dakota MCP Connector
The connector works with any AI assistant that supports MCP. Setup differs slightly by tool, but the underlying access is the same: your AI assistant queries the Dakota MCP server, which returns the data your subscription covers.
→ Claude. Anthropic's AI assistant, available on Pro, Max, Team, and Enterprise plans. Full setup guide: Dakota Marketplace Claude Connector Set Up Guide.
→ ChatGPT. OpenAI's AI assistant, with MCP connector support on Business, Enterprise, and Edu plans. Full setup guide: ChatGPT Set Up Guide.
→ Copilot. Microsoft's AI assistant, with MCP connector support in Microsoft 365 Copilot. Full setup guide: Dakota Marketplace Copilot Studio Early Access Guide.
Additional AI assistants, including Gemini, are compatible with the Dakota MCP server through their own MCP integrations. If your team is already standardized on a specific tool, the connector meets you where you are.
Security and Governance
The Dakota MCP connector inherits the same security model as Dakota Marketplace. Three things worth knowing:
- Access mirrors your existing permissions. Whatever data your Marketplace subscription covers is exactly what the connector returns. If your team subscribes to Dakota FAs but not the full Marketplace, your AI assistant returns FA data only, with no spillover.
- Data stays in your tenant. Authentication happens inside your own environment, and data does not leave it. The connector doesn't route Marketplace records through third-party servers.
- Prompts stay private. Questions asked inside Claude, ChatGPT, Copilot, or any other AI assistant through the Dakota connector are not fed back to Dakota. Your workflow stays your workflow.
The same package-based access controls that govern Dakota API usage carry through to the MCP connector automatically. Nothing extra to configure.
How Dakota Marketplace Fits In
Dakota Marketplace tracks 178,000+ LP Accounts, 435,000+ Verified LP Contacts, 90,000+ Private Funds, 1,240,000+ Private Companies, 31,000+ Transactions.
The MCP connector makes all of it queryable through Claude, ChatGPT, Copilot, or the AI assistant your team already uses, in seconds and in plain English.
Book a demo to see what the connector looks like with your team's workflow and your team's stack.
