I am a versatile professional with a Bachelor’s degree in Cognitive Sciences and Artificial Intelligence from Tilburg University, where I conducted advanced research on avalanche prediction using deep learning techniques, specifically bi-directional LSTM neural networks, improving prediction accuracy for public safety in mountainous regions. My thesis showcased strong skills in data preprocessing, neural network model design, and performance optimization using Python and associated libraries.
In parallel to my academic expertise, I have developed and launched two fully responsive websites for a luxury chalet rental brand in Val d’Isère, built with HTML, CSS, and JavaScript, and one site using Joomla CMS. These platforms feature modern, multilingual designs optimized for high-end client engagement and user experience.
My diverse experiences include working as a ski equipment salesman assisting international customers, an AI engineering intern managing data preprocessing and model selection, and roles in hospitality, municipal services, and logistics. This wide-ranging background helped me develop strong adaptability to various work environments, quickly mastering new skills and technologies to meet organizational goals effectively.
What sets me apart is my ability to bridge technical expertise in AI and web development with practical business and customer-facing experience. This unique blend, coupled with my passion for delivering tailored digital solutions, allows me to confidently adapt and thrive in dynamic, interdisciplinary roles
Experience Level
Language
Work Experience
Education
Qualifications
Industry Experience
This thesis, titled “Avalanche Prediction from Different Nivology and Climatic Parameters Using an RNN,” presents a deep learning approach to improve avalanche danger level forecasting in mountainous winter tourism regions. Current prediction methods rely heavily on subjective human expertise with an accuracy of around 76%. Traditional machine learning models like random forests improve this to about 82% but struggle with handling high-dimensional and temporal data.
The research implements a bi-directional Long Short-Term Memory (BLSTM) recurrent neural network trained on a large dataset of 292,827 samples from the Swiss Alps, covering 20 years of weather and snowpack data. By using sequences of 30 consecutive days as input, the BLSTM model achieved an accuracy of 89%, significantly outperforming human and state-of-the-art machine learning benchmarks. This model captures complex temporal dependencies and processes raw high-dimensional data with minimal preprocessing, making it suitable for real-world applications.
The work contributes scientifically and practically by proposing a more reliable, efficient, and automated avalanche forecasting system aimed at enhancing public safety and supporting winter tourism economies. It highlights the growing importance of data-driven methods to complement expert intuition in environmental risk prediction
A full version of my thesis, detailing the methodology and results of my research on avalanche prediction using recurrent neural networks, is available: https://www.twine.net/signin
Hire a AI Engineer
We have the best ai engineer experts on Twine. Hire a ai engineer today.