Executive Dashboard

Real-time banking risk & retention intelligence powered by advanced ML

Demo Mode: Prediction results are simulated as model hosting has expired. For full functionality with trained models, run locally.

Banking Churn Model

91.6%
Accuracy
95.4% Precision 87.4% Recall
Production Ready

Fraud Detection Model

99.8%
Accuracy
70.9% Precision 75.7% Recall
Business Ready

Dataset Coverage

294K
Total Records
10K Banking 284K Fraud
Complete Coverage

Feature Engineering

64
Engineered Features
Risk Scores Demographics
Advanced Pipeline

Business Impact Analysis

Customer Retention

87.4%

of at-risk customers identified for proactive retention campaigns

Fraud Prevention

99.8%

overall fraud detection accuracy with minimal false positives

Cost Efficiency

95.4%

precision reduces unnecessary retention marketing spend

Risk Coverage

75.7%

of fraudulent transactions caught in real-time

Data Insights & Analytics

Comprehensive analysis of banking customer behavior and risk patterns

Geographic Risk Analysis

Highest Risk: Germany (32.4% churn rate)
Lowest Risk: France (16.2% churn rate)

Age Risk Profile

Peak Risk Age: 50-59 (56.0% churn rate)
Lowest Risk Age: 25-29 (7.1% churn rate)

Gender Risk Distribution

Female Risk: 25.1% churn rate (4,543 customers)
Male Risk: 16.5% churn rate (5,457 customers)

Product Portfolio Risk

Critical Risk: 4 Products (100% churn rate)
Optimal Portfolio: 2 Products (7.6% churn rate)

Customer Value Analysis

Highest Value Risk: Medium Tier (25.2% churn)
Safest Segment: Low Tier (13.8% churn)

Customer Engagement

Inactive Risk: 26.9% churn rate (4,849 customers)
Active Retention: 14.3% churn rate (5,151 customers)

Balance Risk Analysis

Highest Risk: Low Balance (66.7% churn)
Largest Segment: Premium (38.3% customers)

Customer Tenure Impact

New Customer Risk: 22.6% churn (0-2 years)
Most Stable: 18.7% churn (6-8 years)

Strategic Business Insights

Critical Alert

100% churn rate for customers with 4 products. Immediate portfolio optimization needed.

URGENT

High Risk Segment

Female customers 50-59 in Germany represent the highest combined risk profile.

HIGH PRIORITY

Retention Opportunity

2-product customers show 7.6% churn - optimal cross-sell target for single-product users.

OPPORTUNITY

Success Pattern

Active members have 47% lower churn rates than inactive customers.

BEST PRACTICE

Machine Learning Models

Advanced XGBoost models with proven industry benchmarks

Banking Churn Prediction

Accuracy
91.6%
Precision
95.4%
Recall
87.4%
AUC
96.3%

Algorithm: XGBoost with SMOTE

Features: 64 engineered features

Training: 600 estimators, early stopping

Threshold: 0.56 (optimized for recall)

Fraud Detection

Accuracy
99.8%
Precision
70.9%
Recall
75.7%
AUC
94.5%

Algorithm: XGBoost Ensemble

Features: 31 PCA features + Amount_log

Training: 1200 estimators, AUCPR optimization

Threshold: 0.88 (optimized for precision)

Model Performance Curves

Live Predictions

Test our production-ready models with custom inputs

Banking Churn Prediction

Assess customer retention risk using advanced ML models

Customer Profile

300-850 range
Customer age

Geographic & Tenure

Years with bank

Financial Profile

Current balance
Annual income

Banking Products & Activity

Fraud Detection

Real-time transaction fraud risk assessment

Transaction Details

Amount in USD
Seconds from start

Fraud detection uses 28 anonymized PCA features (V1-V28) from the transaction. For demo purposes, we'll simulate typical transaction patterns based on amount and timing.