Portfolio · SQL Projects

Relational SQL
Analytics

Fraud detection, credit risk intelligence and database monitoring using MySQL — turning raw transaction data into actionable risk signals.

← Back to Home
🔍 Fraud Detection System 
-- High-value transactions
SELECT transaction_id, sender_account,
       amount, city
FROM transactions
WHERE amount > 1000000
ORDER BY amount DESC;

SQL · MySQL · Database Views

Fraud Detection Monitoring System

A SQL-based fraud monitoring system built on a simulated banking transaction database. Five analytical business questions are answered to surface suspicious high-value transactions, frequent transfer patterns, high-risk customer behaviour, and geographic transaction hotspots. Four reusable database views enable continuous fraud monitoring.

Key Insight

KES 2.60M in flagged transactions identified — 2 transactions exceeding the KES 1M threshold in Thika and Mombasa. Account 1002 flagged for rapid transfer activity. Esther Achieng (risk score 560) flagged for high-value transaction.

MySQL Fraud Detection SQL Views Joins Aggregations AML
↓ Project Report ↓ SQL File ↓ Query Results
🏦 Bank Risk Intelligence 
-- Branch default exposure
SELECT b.branch_name,
       SUM(l.loan_amount) AS default_exposure
FROM branches b
JOIN loans l ON b.branch_id = l.branch_id
WHERE l.loan_status = 'Default'
GROUP BY b.branch_name
ORDER BY default_exposure DESC;

SQL · MySQL · Credit Risk Analytics

Bank Risk Intelligence System

A relational SQL database project modelling customers, loans, collateral, branches, and transactions to monitor credit exposure, collateral coverage, branch-level default risk, and high-value transaction activity across a simulated banking environment.

Key Insight

James Mutiso, Felix Kiptoo and Linda Chebet hold the highest loan exposure. Kitale and Mombasa branches carry the greatest default burden. Several loans identified as under-collateralised with coverage ratios below 1.0.

MySQL Credit Risk CASE Statements Multi-table Joins Collateral Analysis Views
↓ Project Report ↓ SQL File ↓ Query Results