I'm a 5th-semester Software Engineering student passionate about building backend systems that actually work — clean, scalable, and well-documented. Lately I've been diving deep into AI, exploring computer vision and machine learning beyond the classroom.

• 🔭 Currently working on: deepening my understanding of neural networks & CV
• 🌱 Learning: AI algorithms, deep learning, data science
• 💬 Ask me about: Spring Boot, RESTful APIs, or soybean diseases 🌿
• 🎯 Goal: Land a backend/full-stack internship where I can contribute to real products
• 🌍 Languages: Portuguese (native) · English (Pearson certified)

Deep Learning · Computer Vision · Python · Streamlit

A personal passion project applying AI to agriculture. Trained a ResNet18 model with transfer learning on 3,720 real field images (SoyNet dataset), achieving ~96% accuracy on soybean disease classification. Handled a 6:1 class imbalance with WeightedRandomSampler, exported the model to ONNX, and wrapped it all in a live Streamlit web interface.

PyTorch ResNet18 Transfer Learning ONNX Streamlit Computer Vision EfficientNetV2

Spring Boot · REST API · Gamification · Academic Project (Highlighted by professors)

A corporate platform for structured feedback and continuous improvement, built with Spring Boot across 5 REST modules (feedbacks, tasks, evolution, progress, mentoring). Features a full gamification system with points, badges, and levels, and role-based access for Managers, Collaborators, and Mentors.

Java Spring Boot Spring Data JPA MySQL MVC REST API

Real Client · Spring Boot + Astro · Gmail API · PDF Reports

Built for a real client — massage therapist Valéria de Souza Ávila. Delivered 26 functional requirements including dynamic scheduling, digital patient intake forms, automated Gmail API email notifications, payment tracking with receipt uploads, and PDF financial reports.

Java Spring Boot Astro Gmail API MySQL UML

🤖  Machine Learning & Neural Networks (CS50 AI)
🌾  Computer Vision for Agriculture
☁️  Cloud & Deployment (next step)
🔐  API Security & best practices