AI & MCP

The Payerset Research Assistant

The first Healthcare Price Transparency MCP server live in all four major AI Platforms: Claude, ChatGPT, Microsoft 365 Copilot, and Gemini Enterprise, backed by the most complete dataset in the industry.

  • Quadrillions of records, full rate history
  • Plans, runs, and explains its own analysis
  • Built-in knowledge base for accuracy every step of the way

The Research Assistant

Industry-leading AI on the industry’s most complete dataset

Any AI is only as good as the data behind it. Payerset’s Research Assistant runs on the entire published price transparency corpus, at the most detailed level with no rollups or aggregates, to ensure accuracy.

A full research team, for every user

Guided by Payerset’s knowledge base, it scopes the question, builds the full provider and code universe, pulls every relevant source, and validates before it answers. Every user on your team gets that firepower.

Real rates, never estimates

Every number it surfaces is a real, observed negotiated rate from the source records. It never averages or blends rates into a synthetic figure, and every answer arrives with the evidence behind it.

Every dataset that matters

Transparency data is the foundation. The Research Assistant triangulates it against all-payer claims and remits, payer policies, and dozens of Medicare and Medicaid datasets, so no rate ever stands alone.

Enterprise-grade by default

Per-user OAuth 2.0 on every platform, with your organization’s data rights, entitlements, and usage terms verified on every request. Nothing is indexed into the AI platform, and access can be revoked centrally.

Connect in minutes

One endpoint, four platforms

The same hosted Payerset MCP endpoint serves every platform, with nothing to install locally. Pick your platform for the short version or jump to the full step-by-step guide.

Endpoint https://mcp.tpi.payerset.com/mcp
Claude Anthropic Live · in production Claude Team and Enterprise, plus Claude Code for technical users.
Integration
Organization-installed remote MCP connector
Auth
Per-user OAuth 2.0
Admin
Claude Owner or Primary Owner
Works in
Claude web, Desktop, Cowork, mobile & Claude Code
  1. 1 A Claude Owner adds a custom connector (Organization settings → Connectors → Add → Custom) named Payerset with the Payerset MCP endpoint.
  2. 2 Set organization tool permissions and roll out to a pilot group.
  3. 3 Each user connects their own account (Customize → Connectors → Connect) and completes the Payerset OAuth sign-in.
  4. 4 Enable Payerset in a conversation and start asking.

Technical users can add the same endpoint to Claude Code: claude mcp add --transport http payerset https://mcp.tpi.payerset.com/mcp

Full Claude connector guide
ChatGPT OpenAI Live · in production Direct connections, managed ChatGPT Enterprise rollouts, or your own apps on the Responses API.
Integration
MCP app connection, private workspace plugin, or Responses API
Auth
Per-user OAuth 2.0
Admin
Each user (direct) or workspace admin (managed rollout)
Works in
ChatGPT Enterprise & OpenAI Responses API
  1. 1 Turn on developer mode (Settings → Security and login → Developer mode).
  2. 2 Add Payerset in ChatGPT Plugins with the Payerset MCP URL as a public endpoint.
  3. 3 Enable Payerset in a conversation and complete the OAuth sign-in on first use.

For a managed ChatGPT Enterprise rollout, a workspace admin imports Payerset’s private marketplace package and assigns availability by role. Custom applications connect through the OpenAI Responses API with the same endpoint.

Full ChatGPT connector guide
Microsoft 365 Copilot Microsoft Live · in production Copilot Chat, Copilot in Excel, Researcher, and Cowork through a custom federated connector.
Integration
Customer-managed custom federated connector
Auth
Per-user OAuth 2.0
Admin
Global Administrator or AI Administrator
Works in
Copilot Chat, Copilot in Excel, Researcher & Cowork
  1. 1 Register the Payerset OAuth client in Teams Developer Portal (Tools → OAuth Client Registration) with values delivered during onboarding.
  2. 2 In the Microsoft 365 admin center, create the connector (Copilot → Connectors → Create a new connector) with the Payerset base URL and the OAuth registration ID.
  3. 3 Stage the rollout to a pilot group of licensed users.
  4. 4 Each user adds Payerset as a source in Copilot Chat and signs in.
Full Microsoft 365 Copilot guide
Gemini Enterprise Google Live · in production Gemini Enterprise Standard or Plus through a Custom MCP Server data store.
Integration
Custom MCP Server data store
Auth
Per-user OAuth 2.0
Admin
Discovery Engine Editor + Organization Policy Administrator
Works in
Gemini Enterprise Standard & Plus
  1. 1 Enable the Custom MCP connector policy for the Google Cloud project and allow the Payerset hostnames for egress.
  2. 2 Create a Custom MCP Server data store with the Payerset endpoint and the OAuth values from onboarding.
  3. 3 Load the Payerset actions, enable the approved set, and connect the data store to your Gemini Enterprise app.
  4. 4 Each user enables Payerset actions and signs in with their own account.
Full Gemini Enterprise guide

In practice

Ask bigger questions

The Research Assistant runs each question like a full research engagement: it scopes the market, pulls every relevant source, validates the numbers, and answers with the evidence attached. Work that used to take a week comes back in minutes.

“How do our joint replacement rates compare to every other health system in our market?”

“Why is Anthem paying us so much less for imaging than everyone else?”

“Which payers are underpaying us the most on our highest-volume codes?”

Common questions

Frequently Asked Questions

What is the Payerset Research Assistant™?
The Payerset Research Assistant™ is Payerset’s AI capability, delivered as an MCP server for Claude, ChatGPT, Microsoft 365 Copilot, and Gemini Enterprise. It plans and executes real research on your licensed Payerset data: it scopes the question, pulls every relevant source, validates the results, and presents a defensible answer with the evidence behind it.
How is the Payerset Research Assistant different from a chatbot on a database?
The Payerset Research Assistant™ is different from a chatbot on a database because it carries Payerset’s knowledge base with it: a research methodology built from years of work in this data. The Payerset Research Assistant knows how each payer files, when a number needs validation, and when an absence is itself a finding. Every rate it surfaces is a real, observed number, never a synthetic average.
What data does the Payerset Research Assistant work with?
The Payerset Research Assistant™ works with the entire published price transparency corpus at source-record detail, with full history and no rollups or aggregates. The Payerset Research Assistant triangulates negotiated rates against hospital filings, all-payer claims and remits, payer policies, and dozens of Medicare and Medicaid datasets, all scoped to your Payerset license.
Which AI platforms does the Payerset Research Assistant support?
The Payerset Research Assistant™ supports all four major AI platforms: Claude (Anthropic), ChatGPT (OpenAI), Microsoft 365 Copilot, and Gemini Enterprise (Google). The Payerset Research Assistant is the first healthcare price transparency MCP server available in every one, with customers in production on each, and the same hosted endpoint serves all four.
How is access to the Payerset Research Assistant secured and governed?
Access to the Payerset Research Assistant™ is secured with per-user OAuth 2.0 on every platform. Data is retrieved live at request time rather than indexed into the AI platform, entitlements are enforced server-side on every call, and access can be revoked centrally at any time. Platform administrators can also narrow which Payerset tools the AI may invoke.
How long does it take to connect the Payerset Research Assistant?
Connecting the Payerset Research Assistant™ takes minutes on most platforms: an administrator registers the hosted Payerset endpoint, then each user signs in with their own Payerset account. There is nothing to download or host. Step-by-step guides for each platform are at docs.payerset.com.

Put the Research Assistant to work

Request your demo and watch it answer real questions from your licensed price transparency data in Claude, ChatGPT, Copilot, or Gemini.

  • First MCP server live on all four major AI platforms
  • 100% of published price transparency data at source detail
  • Research and analysis built in, with per-user OAuth

Contact our team