Vithushan Jeyapahan

A versatile and highly motivated software engineer with a strong academic background in Computer Science and Finance. Expert in developing innovative software solutions and machine learning models, with proficiency in Python, Java, C++, and more. Proven ability to lead projects, optimize processes, and collaborate effectively within Agile teams. Passionate about leveraging cutting-edge technologies to solve complex problems and drive business success. Committed to continuous learning and professional growth, ready to tackle diverse software engineering challenges.

Full-Stack Development: Built robust trading platform backend services using Java and Spring Boot, implementing RESTful APIs and microservice architecture for scalable trading operations. Integrated these services with front-end systems to provide real-time trading capabilities. Algorithmic Trading Solution: Developed and managed a semi-automated algorithmic trading solution specifically for financial markets, enhancing market monitoring and risk management, which resulted in a 15% increase in profitability. This involved designing and implementing algorithms that could adapt to market changes, optimise trading strategies, and provide real-time analysis. Machine Learning Integration: Integrated advanced machine learning techniques to improve predictive accuracy by 20%. This included developing models using Python libraries such as TensorFlow and PyTorch and implementing them into the Java-based trading system to provide better forecasting and decision-making support. Process Optimisation: Led the integration of PyTorch and TensorFlow into the Spring Boot workflow, reducing manual processes by 30%. This significantly improved efficiency by automating data pre-processing, model training, and evaluation processes. Collaborative Development: Worked in an Agile environment, collaborating closely with cross-functional teams to ensure seamless integration and deployment of solutions. Utilise version control systems (Git) and CI/CD pipelines to maintain code quality and facilitate continuous improvement. Technological Leadership: Took the initiative to explore and integrate emerging technologies and methodologies to keep the trading solution at the cutting edge, such as incorporating Kubernetes for better orchestration and Docker for containerisation.

Developed an innovative hybrid model integrating sentiment analysis, neural networks, and Bayesian optimization for stock market prediction across multiple sectors, showing strong performance particularly in consumer goods and retail stocks. Algorithm Development: Designed and implemented a hybrid deep learning model combining LSTM networks with GPT-4o Mini-powered sentiment analysis for financial news processing and Bayesian optimization for hyperparameter tuning Multi-Sector Analysis: Successfully tested the model across diverse market sectors, including technology (Tesla, Apple, Meta), consumer goods (Procter & Gamble), and retail (Tesco), demonstrating adaptability to different market conditions Technical Implementation: Built comprehensive data pipeline using Python, TensorFlow, and OpenAI API for sentiment analysis. Implemented web scraping for financial news headlines using BeautifulSoup and automated sentiment scoring system Model Architecture: Created three-layer LSTM network with dropout regularization, incorporating both technical indicators and sentiment scores. Optimized hyperparameters using Bayesian optimization via scikit-optimize Performance Results: Achieved superior prediction accuracy for consumer-facing companies, with notably strong performance in consumer goods and retail sectors. Demonstrated the effectiveness of sentiment integration in price prediction Tools & Technologies: Python, TensorFlow, OpenAI API (GPT-4o Mini), scikit-learn, pandas, numpy, BeautifulSoup, yfinance, matplotlib