Bipsync MCP (Model Context Protocol) Server

Built on the open Model Context Protocol (MCP) standard, it works with any AI assistant that supports MCP — including Claude, ChatGPT, Microsoft Copilot, and Gemini. Bipsync remains your system of record. No data is copied, migrated, or stored elsewhere.

Your unique server endpoint (case sensitive, ensure that the URL is all lowercase):

https://mcp.bipsync.com/{your-client-name}/mcp

Please contact your Bipsync account team to allow Bipsync's MCP server to communicate with your particular client installation and obtain your client URL.

What Can It Do?

Search Your Research

From inside your AI assistant, ask questions and get answers pulled directly from your Bipsync notes and files.

Examples:

"Find my Bipsync notes on [fund manager name]"
"Search for research created in the last week about [topic]"
"Who wrote notes about [company] this quarter?"

The search supports keyword matching, structured filters (sector, analyst, date ranges, custom fields), or both at once. Attachment content (extracted text from PDFs, Word docs, etc.) is also searchable.

Create Notes

Any note you draft in your AI conversation can be saved directly to Bipsync — same audit log, same search indexing, same notifications as notes created in the Bipsync app.

Automatic Context

Behind the scenes, the MCP Server handles additional tasks so your AI assistant can give better answers:

User resolution — when you say "notes by Sarah", it resolves the name to the correct Bipsync user
Entity resolution — when you reference a fund, company, contact, or event by name, it matches to the correct Bipsync record
Identity — the AI knows who you are in Bipsync, so it can filter by "my notes" or scope results to your permissions

Steering (Optional)

If your organisation has Setup App access, a Bipsync administrator can configure a steering document that shapes how AI assistants interact with your Bipsync data across every connected client.

Steering is optional — the MCP Server works fully without it — but it's a powerful way to make responses more consistent and aligned with how your team works.

Steering can guide, for example:

Field naming conventions — the preferred field names and terminology to use
Pipelines and classifications — which pipelines, sectors, or classifications to prioritise or surface first
Response tone and format — how answers should be structured and worded
Scope and defaults — the records and filters most relevant to your team's workflow

Because steering is applied server-side, it takes effect for every user and every AI assistant connecting through your organisation's MCP endpoint, with no per-user setup required. See Getting Started with Bipsync MCP for how administrators configure it.

Supported Fields (Read & Write)

Field Type	Read	Write
Text	✅	✅
Number	✅	✅
Date	✅	✅
HTML	✅	✅
List	✅	✅
Lookup	✅	✅
Group	✅	Coming soon

Supported AI Platforms

Platform	Individual Self-Serve	Enterprise Admin Setup	Write Support
Claude (Free / Pro / Max)	✅ Full self-serve	—	✅
Claude (Team / Enterprise)	✅ Post admin setup	✅ Low effort	✅
ChatGPT (Plus / Pro)	✅ Developer Mode	—	❌ Read-only
ChatGPT (Business / Enterprise)	✅ Post admin setup	✅ Medium effort	✅
Microsoft Copilot (M365)	❌	✅ High effort (Copilot Studio)	✅

The platforms listed above are the ones Bipsync has tested and actively supports. Because the Bipsync MCP Server is built on the open Model Context Protocol standard, any MCP-compatible product or platform can connect to it — this table reflects our verified integrations rather than a hard limit.

If you use another MCP client, it will most likely work. Contact your Bipsync account team if you'd like help getting set up, or if you'd like a specific client verified.