New: Outcome Agents for customer success, services and delivery →

FunnelStory Platform / MCP Server

Centralize Customer Intelligence for Your Favorite AI Copilots.

Empower RevOps, CS Ops, and technical leadership with a direct API connection to the FunnelStory Customer Intelligence Graph. Bring pre-computed, real-time customer data into Claude, ChatGPT, and custom environments while drastically reducing token inflation.

FunnelStory MCP Server architecture: customer intelligence graph connected to AI copilots

Customer teams use FunnelStory to stay ahead of risk and ship intelligence at agentic scale

Secure, On-Demand Access to Your Context Graph.

Connect the context graph to your AI copilot in three simple steps.

1

Authenticate & Connect

Users securely authenticate their FunnelStory identity directly within their chosen copilot (e.g., Claude, ChatGPT).

2

Open Up the Context Graph

The FunnelStory MCP Server gives your copilot governed access to the context graph, bringing every account's data and wisdom directly into the chat.

3

Real-Time, Governed Intelligence

Query data on-demand. The server automatically enforces Role-Based Access Control (RBAC), manages API limits, and streams real-time updates seamlessly.

FunnelStory MCP — claude.ai

What your AI copilot can do

Ask and build from your copilot

Ask anything about your accounts, usage and conversations
Get answers from your product docs, with the source
Build, test and run AI agents
Create tasks and projects for your team
Read your Google Sheets and decks
Connect and model new data

Agents you build can

Read and update Salesforce and HubSpot records
Send email and Slack messages
Read Google Drive docs, sheets and PDFs
Create PDFs and PowerPoint decks
Search the web and enrich contacts and companies
Write results back to BigQuery
Call any API or run custom code
RBAC enforced — scoped to your account permissions

On-Demand Graph Access

Break free from rigid UI limitations. Query the graph from Claude, ChatGPT or any MCP client, get answers from your product docs with the page they came from, and build your own views on the fly.

Automate Scheduled Tasks

Move beyond ad-hoc queries. Use the MCP server to run scheduled, repeatable tasks and Agentic workflows within your Copilot or any 3rd-party orchestration environment.

Pre-Computed Intelligence

Act as a caching layer that delivers pre-computed answers from FunnelStory, avoiding expensive token-based inferences on the client end.

Built-in Wisdom, Not Prompting

A general AI assistant knows nothing about your customers until someone writes it into a prompt. Through the MCP server, it starts from what FunnelStory has already learned about every account: what is true, and what to do about it.

Descriptive wisdom

What's true about the account

  • Stakeholders, champions and what they care about
  • Concerns, use cases and how the relationship has changed
  • What's been remembered from months or years back, with the source

Decision wisdom

What to do about it

  • Which signals actually preceded churn or expansion in your book
  • Which adoption levels renew better among similar accounts
  • Cohorts and next best actions from Outcome Agents

For your post-sales team

CSMs and leaders ask Claude or ChatGPT a question and get an answer that already knows the account, with no context to paste in.

For your post-sales agents

Agents you vibe code start from the same wisdom, so they investigate and act on facts, not on whatever a prompt happened to include.

TOKEN CACHING

Stop Paying for Duplicate Inferences: Eliminate Token Inflation

Feeding raw customer data into LLMs for every single query drains your context window and skyrockets API costs. The FunnelStory MCP Server acts as an intelligent caching layer, delivering mathematically validated, pre-computed intelligence directly to your copilot.

  • Bypass Redundant Compute

    Why ask an LLM to analyze raw usage data when FunnelStory already has the answer? We pass the insight, not just the raw data.

  • Reduce Token Consumption

    Drastically lower your LLM API costs by serving cached, pre-processed context to Claude and ChatGPT.

  • Faster Agent Responses

    Pre-computed intelligence means your agentic workflows execute in a fraction of the time, without hitting rate limits.

The Old Way

Raw dataHuge token payloadSlow and expensive

With FunnelStory

Context graphPre-computedFast and efficient
FunnelStory — Skills Directory

20 Pre-Loaded Skills

Account intelligence

Account briefBook of businessHealth score breakdownSupport history

Meetings

Meeting prep

Risk and renewals

Churn risk reportRenewal dashboardExecutive sponsor coverage

Growth and expansion

Expansion dashboardUpsell dashboardAdoption gapsLead reports

Customer reviews

Internal QBR deckCustomer QBR deckValue emailCase study

Planning, product and marketing

Success planFeature requestsAdoption funnelCustomer ROI stories

Build the Future of Customer Intelligence Operations

Single Source of Truth

Unify your AI toolset. Ensure Claude, ChatGPT, and your custom enterprise apps are all pulling from the exact same mathematically validated signals and context.

Vibe Code AI Agents

Vibe code custom AI agents from Claude, ChatGPT, Gemini or any similar AI platform. Describe the agent in plain language, and your assistant builds it in FunnelStory, tests it on a real account and runs it.

Deploy Pre-Built Skills

20 skills come pre-loaded, from account briefs and meeting prep to churn reports, QBR decks, value emails and case studies. Ask in plain language and the right skill runs on your live FunnelStory data.

Ensured Data Security

Security built for the Enterprise

We've taken the extra steps to ensure your data is safe. Our platform offers Role-Based Access Control and detailed Audit Logging, ensuring your data is protected and monitored at all times

SOC 2 Type IIGDPRISO

Ready to Transform Your Post-Sales?

Unlock the power of the Customer Intelligence Graph