I'm a developer focused on building modern software with clean architecture, thoughtful design, and real-world usability.

I enjoy turning complex workflows into intuitive digital experiences while continuously improving performance, maintainability, and user experience.

• Focused on scalable software architecture
• Passionate about modern UI/UX
• Building clean, maintainable applications
• Always refining one project at a time

Libraries: Recharts • jsPDF • SheetJS (xlsx)

Python Backend: FastAPI • SQLAlchemy • Pydantic • Pydantic Settings • Alembic

Node.js Backend: Node.js • Express.js • Prisma ORM

Authentication: JWT • bcrypt

Caching & Services: Redis

Database: PostgreSQL • Neon

Drivers: psycopg2

Testing: Pytest • HTTPX

Hosting: Vercel • Render

Live Application • GitHub Repository

A real-time forex and gold analysis platform that combines technical analysis, feature engineering, and six locally-trained machine learning models to produce explainable trading recommendations. It does not sell signals or place trades. Instead, it analyzes live and historical market data and provides recommendations you can interrogate, including trend, confidence, risk, expected reward, supporting indicators, and the exact conditions that would invalidate the analysis. When the evidence is not strong enough, it recommends no trade rather than forcing a call.

• Streams live prices for major forex pairs and XAU/USD
• Stores historical market data and displays real-time candlestick charts
• Computes technical indicators and price-action features in-house
• Uses RSI, MACD, ADX, Bollinger Bands, market structure, liquidity sweeps, and fair value gaps
• Runs six locally-trained ML models for trend, entry quality, confidence, risk, reward, and market regime
• Combines model outputs into one explainable recommendation
• Clearly identifies the indicators and conditions supporting each recommendation
• Provides explicit invalidation conditions for every analysis
• Recommends no trade when the available evidence is insufficient
• Supports backtesting across multiple trading strategies
• Includes trend-following, SMC/liquidity, range, breakout, swing, and scalping strategies
• Provides a paper trading environment with virtual portfolios
• Supports stop-loss/take-profit execution and trade journaling
• Includes registration, email verification, password reset, and account deletion
• Uses JWT authentication with rotating refresh tokens

Frontend: React 19 • TypeScript • Vite • Material UI • TradingView Lightweight Charts • Recharts

Backend: Python • FastAPI • SQLAlchemy • Alembic

Database: PostgreSQL • TimescaleDB • Neon

AI/ML: scikit-learn • XGBoost • LightGBM • pandas • NumPy

Real-Time: WebSockets • Redis Pub/Sub • Upstash

Hosting: Vercel • Northflank

Live Application • GitHub Repository

A modern business management platform designed to help small and medium-sized businesses streamline inventory, sales, customer management, supplier operations, reporting, and point-of-sale workflows through a fast, intuitive, and scalable web application.

• Modern dashboard with real-time business insights
• Inventory and stock management
• Supplier management and purchasing workflows
• Sales and customer management
• Point of Sale (POS)
• Business reporting and analytics
• Secure JWT-based authentication
• Responsive Material UI interface
• Centralized business operations and data management
• Designed with scalability and maintainability in mind

Frontend: React 19 • TypeScript • Vite • Material UI • MUI DataGrid • Recharts • jsPDF • SheetJS

Backend: Node.js • Express.js • Prisma ORM

Database: PostgreSQL • Neon

Authentication: JWT • bcrypt

Hosting: Vercel • Render

© 2026–Present KAL