Every dealmaker has been there: you’re 10 minutes from a founder call, you’ve got Claude open, and you’re asking it to help you prep. The problem is, it has no idea who the company is, what stage the deal is at, or that your colleague had coffee with the CEO last month. So you open another tab, pull up Affinity, copy-paste the relevant context, and start over.
The Affinity MCP Server is built to eliminate that gap. Now in beta, it’s an open-source server using the Model Context Protocol (MCP) that connects Affinity directly to the AI tools your team already uses, including Claude, Gemini, GitHub Copilot, and others. Your relationships, contacts, deal pipeline, notes, and meeting history are accessible right inside your AI workflows, without switching tabs or reconstructing context from scratch.
Why we built this
Dealmakers are increasingly using AI tools for daily workflows like meeting prep, company research, note-taking, and pipeline analysis. But until now, those tools had no visibility into Affinity, which is where your actual deal data lives. That meant either manually transferring information from your CRM into your AI tool, or making decisions based on whatever the AI could infer from public sources alone.
The Affinity MCP Server fixes this by translating natural-language questions into real-time API calls against your live Affinity data. When you ask your AI assistant to help prepare for a meeting, it’s pulling directly from your source of truth.
Most AI tools are like a brilliant research assistant who's never been inside your office. They can tell you everything about a founder from public sources, but they've never read your notes, seen your team's emails with her, or know she's been sitting in your pipeline for six months. The MCP server gives that assistant a full set of keys, so when you ask about a company, they're not guessing from the hallway.
What you can do today
The beta supports 20+ operations across three workflows that cover the moments dealmakers feel the most friction.
Search your CRM conversationally. Instead of navigating list views and filters, ask in plain language:
- “Which fintech companies have we met with this year?”
- “Search for companies in the healthcare industry with more than 50 employees.”
- “What notes did I create last week?”
The answers come back in seconds, directly inside your AI tool, so you stay in the flow of work rather than breaking to look something up.
Prepare for any meeting in seconds. Meeting prep that used to mean 20 minutes across multiple tabs now takes a single prompt:
- “Help me prepare for my meeting with [Portfolio Company]. Give me their deal status, recent notes, and key contacts.”
- “Summarize the status of deals in our pipeline.”
The AI combines what’s in Affinity with publicly available information, so you’re walking into every call with both your internal context and the full external picture.
Update records without leaving your AI tool. Capturing notes after a call is faster when you don’t have to navigate back to Affinity:
- “Create a note attached to our open [Company Name] deal with today’s takeaways.”
Fewer context switches means better data hygiene, since notes actually get written while the meeting is still fresh.
What’s coming next
This is a beta, and the roadmap reflects what we’re hearing from firms actively using it. On deck: field and status updates on list items, call transcript search, relationship scoring, and more advanced write operations. There’s even a built-in /feedback command so you can flag what’s working or missing directly from inside your AI tool. Your input shapes what gets prioritized next.
{{request-demo-a="/rt-components"}}
Get Started
We’re excited to put this in your hands. If you’re ready to bring your relationship data into your AI workflows, book a demo and our team will walk you through what’s possible with AI-powered relationship intelligence.
The future of CRM is conversational. With the Affinity MCP Server, your most important relationships are now just a question away.
.png){kind=logo}
