DV

Dragos Vrabie

Computer Engineering Student

@ Polytechnic University of Bucharest

About Me

Passionate and dedicated Automation and Computer Science student with proven leadership experience in student organizations such as LSAC and BEST. Skilled in organizing events, leading teams, and fostering collaboration across multicultural environments. Experienced in web development, robotics, software development and automated testing, with a strong commitment to continuous learning and personal growth. In my personal time, I like to take photos, travel and play sports.

Skills

Technical Skills

C/C++ Python Java JavaScript HTML/CSS Arduino Raspberry Pi Git Linux

Languages & Soft Skills

Romanian

C2

English

C2

German

A2

Teamwork Time Management Effective Communication Critical Thinking Photography

Projects

Investment Calculator with Analytics

Comprehensive Java investment calculator featuring compound interest calculations, interactive growth charts, pie chart breakdowns, and detailed financial schedules. Supports multiple compounding frequencies, periodic contributions/withdrawals, and includes automated testing with CI/CD pipeline.

Java 11+ Swing GUI JFreeChart BigDecimal Precision JUnit Testing GitHub Actions Maven Financial Analytics

View Details

Investment Spear

Always-on Python trading system using Yahoo Finance news, local Ollama sentiment analysis, and Alpaca paper trading. Includes a Flask dashboard and conservative cash-only risk management.

Python 3.11+ Flask Ollama Alpaca yfinance SQLite Pytest

View Details

Contact

dragosvrabie@proton.me

Bucharest, Romania

© 2026 Dragos Vrabie. All rights reserved.

Investment Calculator with Advanced Analytics

A comprehensive Java Swing application for calculating compound interest investments with interactive charts, detailed schedules, multi-currency support, and automated testing infrastructure.

Project Overview

This investment calculator provides users with powerful tools to plan their financial future. It features a modern GUI built with Java Swing and FlatLaf look-and-feel, offering comprehensive investment calculations with visual representations and detailed breakdowns. The application supports various compounding frequencies, contribution schedules, negative contributions (withdrawals), and includes a robust testing suite with CI/CD pipeline.

Key Features

Compound interest calculations with multiple frequencies (daily, weekly, monthly, quarterly, annually)
Interactive investment growth charts and pie chart breakdowns showing final amount composition
Support for negative contributions (withdrawals) to model money taken out of investments
Support for regular contributions with flexible timing (beginning/end of period)
Multi-currency support (USD, EUR, GBP, JPY, CAD, AUD)
Comprehensive annual and monthly investment schedules with dynamic labeling
Chart export functionality (PNG/JPEG) with full-screen viewing mode
Automated test suite with detailed input/output validation
GitHub Actions CI/CD pipeline for continuous integration
Comprehensive documentation including calculation methodology

Technical Implementation

The application uses a modular architecture with separate components for calculations, charting, and UI. It implements the Model-View-Controller pattern with dedicated classes for investment engine calculations, chart generation using JFreeChart, and a clean Swing-based user interface. The calculation engine uses BigDecimal arithmetic for financial precision and includes comprehensive validation against known mathematical formulas.

Technologies Used

Java 11+ Java Swing FlatLaf Look and Feel JFreeChart BigDecimal Precision Maven Build System JUnit 5 Testing GitHub Actions CI/CD MVC Architecture

Advanced Features

Dual Chart System: Growth over time charts and pie charts showing investment breakdown (starting amount, contributions, interest earned)
Withdrawal Support: Handle negative contributions with appropriate UI feedback and calculations
Precision Calculations: Uses BigDecimal with 10 decimal places to avoid rounding errors in financial calculations
Comprehensive Testing: Automated test suite validating calculations against known formulas with detailed input/expected/actual output
CI/CD Pipeline: GitHub Actions workflow for automated testing on every push and pull request
Documentation: Detailed explanation of calculation differences between investment calculators and methodology transparency
Export Capabilities: Save high-resolution chart images for reports and presentations
Professional Architecture: Clean separation of concerns with testable, maintainable code structure

Quality Assurance

Automated Testing: Comprehensive JUnit test suite covering edge cases and mathematical validation
Continuous Integration: GitHub Actions pipeline ensuring code quality on every commit
Calculation Transparency: Detailed documentation explaining methodology and potential discrepancies with other calculators
Input Validation: Robust error handling and user input validation with clear error messages

View GitHub Repository

Investment Spear

An always-on Python trading system that monitors Yahoo Finance news, uses a local Ollama model for sentiment analysis, and places Alpaca paper trades with conservative, cash-only risk controls. It also includes a web dashboard for viewing portfolio activity, news, and AI-driven decisions.

Project Overview

Investment Spear is designed to react to changing market conditions instead of trading a fixed list of stocks. On startup, it uses AI market research to build a diversified symbol universe, then refreshes that universe daily while keeping any currently held positions in scope. The bot continuously pulls market news, analyzes it with local AI, and makes buy, sell, or hold decisions based on sentiment, portfolio context, and risk rules. A Flask-based dashboard provides visibility into the portfolio, news flow, and decision history.

Key Features

AI-driven symbol selection and daily universe refresh
News-based trading decisions using Yahoo Finance headlines
BUY / SELL / HOLD logic with portfolio context and sentiment analysis
Alpaca paper trading with cash-only execution
Conservative risk management with reserve capital
Live dashboard for portfolio, news, and decision tracking
SQLite logging for trades and decisions
Automated test suite for core trading behavior

Technical Implementation

The application is built around a continuous trading loop that coordinates market research, news ingestion, decision-making, and execution. A local Ollama model handles symbol research and sentiment analysis. Yahoo Finance provides the news input, Alpaca handles trade execution, and SQLite stores the bot’s activity. The dashboard exposes portfolio and analysis data through a Flask API and a browser UI. The architecture is modular, separating AI logic, broker access, data fetching, storage, and the live bot loop.

Technologies Used

Python 3.11+ Ollama Alpaca API Yahoo Finance / yfinance Flask SQLite Pytest HTML/CSS/JS Frontend

Advanced Features

Daily refresh of the tracked symbol universe
Preservation of held positions during universe rotation
Batch news scanning across the full symbol set
Cash-only trading with no margin usage
Local AI execution for privacy and lower external dependency
Decision logging for later review and debugging
Separate dashboard for monitoring live activity
Automated tests covering bot behavior and supporting services

Quality Assurance

Regression tests for symbol universe merging and news batch processing
Mocked tests for broker, data, and strategy logic
Clear separation of concerns across bot, AI, broker, and data layers
Conservative defaults to reduce trading risk
Documentation for setup, runtime behavior, and bot options

Read Guide View GitHub Repository