Overview

Churn Predictor Ai MCP provides AI-powered tools via the Model Context Protocol (MCP).

Tools

Tool	Description
predict_churn	Predict churn risk for a customer
batch_predict_churn	Predict churn for multiple customers
get_customer_risk	Get current risk score for a customer
track_customer	Add customer to churn tracking
update_customer_signals	Update customer behavioral signals
get_at_risk_customers	Get all customers above risk threshold
create_retention_action	Create a retention action for at-risk customer
get_retention_actions	Get retention actions for a customer
update_retention_action	Update status of retention action
get_churn_analytics	Get overall churn analytics
get_cohort_retention	Get cohort retention analysis
simulate_intervention	Simulate impact of retention action
get_risk_trends	Get risk score trends over time

Installation

pip install meok-churn-predictor-ai-mcp

Usage with Claude Desktop

Add to your Claude Desktop MCP config (claude_desktop_config.json):

{
  "mcpServers": {
    "churn-predictor-ai": {
      "command": "python",
      "args": ["-m", "meok_churn_predictor_ai_mcp.server"]
    }
  }
}

Usage with FastMCP

from mcp.server.fastmcp import FastMCP

# This server exposes 13 tool(s) via MCP
# See server.py for full implementation

License

MIT © MEOK AI Labs