Data Scientist @ Sirros IoT (September 2021 - Present)
Web Developer Internship @ Sirros IoT (November 2020 - September 2021)
Technician in Computer Science - Technological and Communication Axis | Synodal Center of Novo Hamburgo Education Unit of the Evangelical Foundation (January 2020 - November 2021)
Luna, the first AI-powered fashion assistant that offers personalized clothing recommendations with images and direct store links. Combining my love for fashion and expertise in NLP, Luna is my way of reimagining the shopping experience.
I built a machine learning model that doesn’t just predict loan defaults—it explains why. By combining XGBoost with SHAP, I tackled challenges like class imbalance and made the model’s decisions more transparent and trustworthy.
The result? A system that could help lenders make smarter decisions, reduce defaults, and save millions 💵.
Used The Quick, Draw! Dataset from Google to build a CNN to predict which object the person drew. Using the webcam, our hands are the cursor to play the game, within the objective to make the Neural Network to guest which object you drew, given a certain object by the system. The user can select a pen, rubber among other things to draw in the screen.
Used Tensorflow and OpenCV to build a CNN and an algorithm that can detect people’s emotions through facial expressions.