Treasury Management System
A streamlined treasury management system demonstrating SQL-first architecture with core cash management, FX risk tracking, and reporting capabilities powered by Supabase views and stored functions.
Project Purpose
This portfolio project demonstrates my ability to architect SQL-first applications with database views and stored functions at the core. The Treasury Management System showcases capabilities in PostgreSQL optimization, FX risk calculations, cash position aggregation, and financial data processing.
Note: This is a demonstration project using fictional data. All financial figures, transactions, and forecasts are simulated for portfolio purposes.
Business Problem Solved
Treasury departments struggle with fragmented cash and FX data, leading to:
❌ Without Centralized Treasury
- • Manual cash position calculations
- • Spreadsheet-based FX tracking
- • No real-time visibility
- • Time-consuming reconciliation
- • Error-prone reporting
- • Disconnected bank accounts
✅ With This System
- • Automated cash aggregation
- • Real-time FX exposure tracking
- • Multi-currency dashboard
- • Instant cash forecasting
- • SQL-powered calculations
- • Centralized treasury view
Project Metrics
Target Users
Treasury Managers
Monitor cash positions, manage FX exposure, and optimize currency hedging strategies
Finance Teams
Track cash forecasts, analyze FX P&L, and generate treasury reports for stakeholders
CFOs
View consolidated cash positions, assess FX risk, and make informed treasury decisions
What Makes This Project Stand Out
Ensemble ML Architecture
Combines Linear Regression, Random Forest, and Moving Average models for superior prediction accuracy through weighted ensemble approach.
Multi-Method Anomaly Detection
Uses Z-score, IQR, and Isolation Forest algorithms to identify outliers with severity classification and natural language explanations.
Feature Engineering Pipeline
Automated feature creation including time-based, lag, rolling statistics, and growth features for enhanced model performance.
Real-Time Processing
FastAPI backend with async processing, indexed PostgreSQL queries, and efficient data aggregation for sub-second response times.
Ready to See It In Action?
Explore the complete cash flow prediction system with ensemble ML models, real-time anomaly detection, and interactive financial dashboards.
💡 Tip: Click "Generate Forecast" to see ML predictions in real-time!
Skills Demonstrated in This Project
This is a portfolio demonstration project using fictional data. Built to showcase advanced full-stack ML development capabilities.
© 2025 Portfolio Project • All Rights Reserved