A selection of my work in Data Science, AI Agents, and Engineering.
A comprehensive full-stack web application built with Flask and HTML to track my stock portfolio. Features real-time data updates, Discounted Cash Flow (DCF) analysis, a wishlist for potential investments, a page for adding notes on stocks, and performance metrics.
An autonomous Multi-Agent System to research financial markets. Agents scrape news, analyze sentiment, check technical indicators, and compile reports. The system automatically generates a comprehensive stock research report for any ticker symbol entered.
The website you are looking at now — minimalist, high-performance and responsive. Pages: Home, Projects, CV, and Blog. I built this website from scratch because I wanted full control over the design and functionality. It's built with Flask for easy content management and deployment.
For my Bachelor Thesis, I conducted a detailed investigation into awe and the Overview Effect, using the CAVE VR system at the DAF Technology lab. My research focused on how virtual reality experiences can induce feelings of awe and the Overview Effect.
A project focused on exploring and analyzing EEG data using the MNE library in Python. Includes data preprocessing, visualization of brain activity, and extraction of meaningful insights from EEG signals.
A small hands-on project to run OWASP Juice Shop (an intentionally vulnerable web app) and OWASP ZAP (free proxy/scanner) to learn basic web security concepts.
This project aimed to explore historical S&P 500 data and build machine learning models (An LSTM model and a LightGBM model) to identify patterns related to bull markets, bear markets, and crashes, investigating potential predictability in market movements.
A deep learning project focused on predicting valence (emotional positivity or negativity) from raw speech data using a bidirectional LSTM model. The project involves preprocessing audio data, feature extraction, model training, and evaluation.
A Tilburg University project for the course 'Autonomous Systems', involving the improvement of an autonomous robot agent designed to escape from a pit obstacle in a simulated environment (BB8 robot). The focus is on robot navigation, obstacle avoidance, and autonomous decision-making using Python.
A Tilburg University project for the course 'AI For Nature & Environment'. Main goal: Construct a predictive neural network model designed to analyze random occurrences spanning diverse locations in Australia.
A small-scale project using various machine learning techniques. The project involves developing a classification model to predict the likelihood of heart attacks based on various health parameters.
This project implements a complete data science workflow using 2.5 years of personal workout data from StrengthLog. It demonstrates the full pipeline from raw data ingestion through feature engineering to predictive modeling, with emphasis on code quality, modularity, and testing.
A lightweight stock analysis tool built in Python (Flask) that helps me perform DCF analyses on stocks quickly. Moreover, I can add stocks to my wishlist with real up-to-date data fetched from APIs. Finally, I can add due diligence notes for each stock.
A simple workout logging web application built with Flask and PostgreSQL. It allows users to log their workouts, view PRs, log their weight, and track progress over time.