Machine Learning & AI

• NeuroStorm: Predictive model linking geomagnetic storm data with stroke risk, developed in a cross-functional team. Presented via interactive dashboard and awarded 1st place in the CASSINI Hackathon (Germany 2025)
• DrIDENT: App for pill identification (Transfer Learning model; https://github.com/nitro381penta/pill-recognition-app)
• CiteRAG: Local-first Retrieval-Augmented Generation (RAG) system for PDF-only corpora designed for research and analysis (https://github.com/nitro381penta/CiteRAG)
• BioRaG: RAG system to answer questions on topics related to the use of Machine Learning and AI in biology
• Automated Customer Reviews: NLP-Model to automate the processing of customer feedback for a retail company by classifying reviews as ‚Positive‘, ‚Neutral‘ or ‚Negative‘
• Bone Fracture Classifier: Convolutional Neural Network to classify bones as fractured or not fractured
• Image Classifier: Convolutional Neural Network for image classification
• Sales prediction model: Model to predict the sales of retail stores

Virtual Reality

• VR CHARLIE: Chat Assistant for Digital OSCE Examinations, VTT Technology Publications 2025, No. T440, p. 221. Available at: https://publications.vtt.fi/pdf/technology/2025/T440.pdf  (Awarded Best Poster at EuroXR 2025)
• SYNAESTHESIA: A VR sound experience with spatial audio and music visualizations (https://github.com/nitro381penta/SYNAESTHESIA)
• A Guide Beyond Space: A VR museum experience exploring art through spatial interaction, immersive perception, and conversational AI (https://github.com/nitro381penta/a-guide-beyond-space)

Tools

• GLYPH registry: A lightweight command-line utility written in modern C++ for the structured registration, organization, and analysis of digital assets