Software Developer

A Bitwarden/Keeper-style password manager implementing Apple's Keychain sync architecture patterns. Built with security-first principles and cross-platform support.

Key Features:

• Triple-layer data model with keys, inet, and ckmirror tables
• Layered encryption using AES-256-GCM with PBKDF2 (100k iterations)
• Conflict-free synchronization with generation counters
• Trusted peer circle using Ed25519 cryptographic verification
• Cross-platform: Desktop (Electron/Angular), Mobile (Flutter), Server (Go)
• Multi-tenant PostgreSQL or single-user SQLite modes

Tech Stack:

Backend: Go 1.24+, PostgreSQL, SQLite, Gin, WebSocket

Desktop: Electron, Angular, TypeScript, SCSS

Mobile: Flutter 3.32.5+, Dart

Security: PBKDF2, AES-256-GCM, Ed25519, HKDF

Challenge: Implementing Apple's sophisticated Keychain sync patterns while maintaining compatibility across multiple platforms and databases.

Enterprise-grade social travel platform combining real-time flight tracking with social networking capabilities. Built for scalability and performance.

Key Features:

• Automated boarding pass OCR scanning with 98% accuracy
• Real-time flight status tracking with push notifications
• Social timeline with photo/video sharing capabilities
• Travel companion matching algorithm
• Location-based check-ins and discovery
• Cross-platform PWA, iOS (12.0+), and Android (API 23+)

Tech Stack:

Frontend: React 18, TypeScript 5.0, Vite 5.0, Tailwind CSS, Apollo Client

Backend: Node.js 18 LTS, Express.js, GraphQL, Apollo Server, Socket.io

Database: MongoDB 6.0, Mongoose ODM, Redis

Mobile: Flutter 3.0 with Provider State Management

Performance:

• Lighthouse Score: 95+
• First Contentful Paint: <1.5s
• API Response Time (p99): <100ms
• Tested with 10,000+ concurrent users

Smooth animated eyes project for ESP32-S3 microcontroller with TFT display. Features interactive behaviors and customizable animations.

Key Features:

• Smooth blinking and pupil tracking animations
• Adjustable eye spacing and dimensions
• Optional eye tracking mode (via serial input)
• Winking, random blinking, and mood-based behaviors
• Dynamic color shifting based on mood states
• Optimized for 60 FPS performance on embedded hardware

Hardware:

• LILYGO T-Display-S3 ESP32-S3 development board
• Built-in 1.9-inch ST7789 TFT display (320x170 resolution)

Tech Stack:

Language: C++ (100%)

Display: TFT_eSPI library

Platform: Arduino IDE compatible, ESP32-S3

Customization: All animation parameters are easily configurable through code variables including frame delay, blink duration, and eye dimensions.

Educational machine learning project for predicting stock price movements using historical data and technical analysis. Implements a complete ML pipeline from data collection to model evaluation.

Key Features:

• 60+ technical indicators and features (MA, RSI, MACD, Bollinger Bands, volume analysis)
• Multiple prediction targets: directional movement, percentage change, absolute price
• Multiple ML models: Random Forest, Logistic Regression, SVM
• Time series cross-validation (proper temporal splitting)
• Interactive Jupyter notebooks for exploration

Tech Stack:

Data: yfinance, pandas, numpy

ML: scikit-learn (classification & regression)

Analysis: ta (technical indicators)

Visualization: matplotlib, seaborn, Jupyter

Results:

Typical accuracy: 55-58% on daily direction predictions. This project demonstrates that stock prediction is inherently difficult and results should be used for educational purposes only.

Note: This is an educational project. Not financial advice. Stock markets are unpredictable and past performance doesn't guarantee future results.