• 🧠 Recently completed a bootcamp in Data Science and AI at Le Wagon
• 📜 Certified Azure Data Scientist DP-100
• 🎓 Background in Mechanical Engineering
• 💼 Experience in Data Analytics and Dashboards as Production Manager in a MedTech startup

• 📽️ SignLens - Sign Language Translation powered by AI

  • For our final project of Le Wagon BootCamp, we created an application to translate sign language videos into written words
  • We built a recurrent neural network (RNN) model on a dataset of 94k preprocessed videos using Mediapipe
  • The model's accuracy reached 61.3% out of 250 different signs, with a test set size of 20%
  • We implemented an API deployed on Google Cloud Platform and built a user-friendly interface using Streamlit. This interface includes features such as video translation, a learning platform, and an engaging practice game
  • Explore the project repositories for further details: deep learning repo and web interface repo

• Dashboards:

  • Geneva Transport Analysis: A dashboard analyzing public transportation data in Geneva using open datasets.
    Technologies: Plotly, Jupyter-flex

  • Lichens Go (Ongoing): A dashboard tracking pollution in urban environments, based on lichen observations on trees.
    Technologies: Plotly, Dash, Dash Mantine Components

• 🏆 2nd place at NovHack 2024

  • Secured 2nd place in a 2-day Data Science Hackathon organized by Euranova, focused on "Reinforcement Learning in the Energy Sector"
  • The objective was to develop an Active Network Management (ANM) model for electricity distribution
  • Using an A2C model (Synchronous Advantage Actor Critic) with MultiInputPolicy, we strategized actions concerning both active (P) and reactive (Q) power injection for individual renewable generators and distributed energy storage units. The primary objective was to minimize energy wastage and prevent breaches of network constraints, such as cable capacity limits

📧 Feel free to get in touch! Let's connect and collaborate on exciting data projects! 🚀