Based in The Netherlands
Transforming raw data into deployable intelligence. I build systems spanning machine learning, data engineering, and full-stack development.
Transforming raw data into deployable intelligence. I build systems spanning machine learning, data engineering, and full-stack development.
I'm Jonas, a third-year Data Science & AI student at Breda University of Applied Sciences. I'm passionate about building intelligent systems that solve real-world problems, from computer vision pipelines to full-stack AI platforms.
Across 10+ projects I've worked with deep learning, NLP, reinforcement learning, computer vision, and full-stack development. I thrive at the intersection of research and engineering, turning raw data into deployable, impactful solutions.
When I'm not training models or wrangling data, you'll find me producing electronic music, diving into the latest AI research papers, or cycling through the North Brabant countryside. I thrive in teams that value curiosity and aren't afraid to experiment.
Statistical modeling, deep learning, NLP, and computer vision. From BERTje to U-Net to reinforcement learning agents.
ETL pipelines, REST APIs, Docker deployments, and cloud infrastructure on Azure. Building systems that scale.
Full-stack AI platforms: from LLM-powered recruitment tools to smart campus analytics. End-to-end from prototype to production.
Predictive analytics system forecasting campus occupancy at BUas. Time series ML models integrating data from cameras, WiFi, and scheduling systems. REST API with real-time endpoints feeding dashboards and automated staff planning.
Predictive occupancy forecasting system for BUas campus operations. Ingests data from entrance cameras, WiFi, TimeEdit room bookings, KNMI weather, and NS train disruptions into a 112-column hourly feature table. Trains a Temporal Fusion Transformer alongside LightGBM, XGBoost, and CatBoost models to forecast building occupancy up to 7 days ahead, feeding FastAPI endpoints and Power BI dashboards used by catering, facilities, and cleaning teams.
Web app that automatically clusters a Spotify library into vibe-based playlists using Last.fm genre tags, MiniLM semantic embeddings, and optional Spotify audio features. Runs hourly, routes new tracks via cosine similarity, and features a real-time SSE-powered dashboard.
AI-powered recruitment platform for Dutch businesses. Uses GPT-4o for bias detection in job postings, CV parsing, ESCO-based skill matching, and EU AI Act compliance. Includes an ML Judge framework with 80 test cases for systematic prompt evaluation.
Deep learning image segmentation pipeline analyzing plant root images using a U-Net CNN. Features CLI, web interface, and REST API, deployed both on-premise and on Azure Container Apps with full CI/CD.
Computer vision + robotics pipeline for the Netherlands Plant Eco-phenotyping Centre. Root segmentation (F1: 0.85), root system architecture extraction, and reinforcement learning agents controlling a liquid handling robot.
NLP pipeline for Dutch language emotion classification. Transcribes video/audio via Whisper, classifies 7 emotions using fine-tuned BERTje/RoBERTa. Achieved 84.7% accuracy with SMOTE-balanced datasets.
Machine learning system predicting road accident risk levels (low / medium / high) for Breda. Merges the ANWB Safe Driving Dataset with six years of Breda accident records, trains ensemble and neural-network models, and deploys via Streamlit.
Deep learning image classifier that identifies 7 global biomes from photos. Built through 4 iterative model experiments culminating in a VGG16 transfer-learning model (~90% test accuracy), with GradCAM + LIME interpretability analysis and a user-tested Figma mobile app prototype.
End-to-end data analysis and ML pipeline for NAC Breda's player recruitment. Cleaned and explored a dataset of 16,535 footballers across 140 features, segmented players by position, and built a Logistic Regression classifier to identify high-potential forward prospects.
Interactive dashboard investigating UN SDG progress, focusing on SDG 14 (Life Below Water), examining how marine ecosystem protection influences fish stock populations.
Academic research paper investigating how personal data security influences customer trust in AI chatbot interactions. Combines qualitative thematic analysis with quantitative survey data, correlation and regression analysis.
Building predictive occupancy system for BUas: time series forecasting, APIs, and automated staff planning.
Breda University of Applied Sciences · 3rd year, 10+ projects spanning ML, NLP, CV, and full-stack development.
Whether you have a data challenge, a collaboration idea, or just want to connect, I'd love to hear from you.
or reach me directly at Jonasvos01@gmail.com