Churn Predictor AI MCP Server

By MEOK AI Labs — Customer churn prediction, retention analytics, and intervention simulation

Installation

pip install churn-predictor-ai-mcp

Usage

# Run standalone
python server.py

# Or via MCP
mcp install churn-predictor-ai-mcp

Tools

predict_churn

Predict churn risk for a customer based on behavioral signals.

Parameters:

• customer_id (str): Customer identifier
• last_login_days (int): Days since last login
• support_tickets_last_30d (int): Support tickets in last 30 days
• nps_score (int): NPS score (1-10)
• usage_decline_percent (int): Usage decline percentage
• payment_failures (int): Number of payment failures
• tenure_months (int): Customer tenure in months

batch_predict_churn

Predict churn for multiple customers at once.

Parameters:

• customers (list): List of customer data dicts

get_customer_risk

Get current risk score for a tracked customer.

Parameters:

• customer_id (str): Customer identifier

track_customer

Add customer to churn tracking system.

Parameters:

• customer_id (str): Customer identifier
• initial_data (dict): Initial behavioral data

update_customer_signals

Update customer behavioral signals and recalculate risk.

Parameters:

• customer_id (str): Customer identifier
• signals (dict): Updated signal data

get_at_risk_customers

Get all customers above a risk score threshold.

Parameters:

• threshold (int): Risk score threshold (default 60)
• limit (int): Max results (default 50)

create_retention_action

Create a retention action for an at-risk customer.

Parameters:

• customer_id (str): Customer identifier
• action_type (str): Action type (outreach, discount, upgrade, etc.)
• description (str): Action description

get_retention_actions

Get retention actions filtered by customer or status.

Parameters:

• customer_id (str): Filter by customer
• status (str): Filter by status

update_retention_action

Update status of a retention action.

Parameters:

• action_id (str): Action identifier
• status (str): New status

get_churn_analytics

Get overall churn analytics for a period (7d, 30d, 90d).

Parameters:

• period (str): Time period (default '30d')

get_cohort_retention

Get cohort retention analysis.

Parameters:

• cohort_month (str): Cohort month

simulate_intervention

Simulate the impact of a retention action on customer risk score.

Parameters:

• customer_id (str): Customer identifier
• action_type (str): Action type (outreach, discount, upgrade, survey, bonus)

get_risk_trends

Get risk score trends over time for a customer.

Parameters:

• customer_id (str): Customer identifier
• days (int): Number of days to look back (default 30)

Authentication

Free tier: 15 calls/day. Upgrade at meok.ai/pricing for unlimited access.

License

MIT — MEOK AI Labs