🌟 My Journey
At the age of 13, I had the opportunity to take robotics classes, and that was the moment one of the greatest passions of my life was awakened. Working with logical reasoning, understanding how machines and programs around me functioned, and learning how these devices were created were activities that fascinated me from a very young age.
That spark was lit the instant I made my first robot move (it only went in a straight line), and it became the first step toward the woman I’ve become today.
At 18, still following that dream, I started my undergraduate degree in Information Systems at UFMG. But I wanted more. I joined my first undergraduate research project — on robot swarms — where block-based programming turned into Python and ROS. In 2019, I decided to take a major leap and pursue two degrees at the same time, also enrolling in Control and Automation Engineering at PUC Minas. Through this path, I discovered Machine Learning and Computer Vision, especially within industrial and medical contexts — fields that have continued to inspire me since.
During my academic and professional journey, I’ve had the opportunity to work in different areas of AI. At Accenture, I worked with Machine Learning models applied to business problems. At TCS Industrial, I focused on Computer Vision solutions for industrial automation.
Today, I’m a Master’s student in Artificial Intelligence at UFMG, advised by Professor Adriano Veloso, and I work as a Data Scientist at Kunumi, where I combine academic research with practical AI applications. My current research focuses on generative AI and reinforcement learning in large language models (LLMs) — exploring how we can teach models to generate structured, controllable, and semantically meaningful outputs, even in low-data settings.
There is still much to learn, more problems to solve, and more systems to build — and I’m truly grateful for that.
💼 Technical Focus
My core expertise is in:
- Generative AI and Reinforcement Learning for LLMs
- Training models to generate structured outputs and synthetic data
- Designing reward functions and evaluation strategies for fine-tuning LLMs
- Experiment tracking and optimization with Weights & Biases and SkyPilot
I also have experience with image-based projects, such as a Brain Age Classifier using MRI data to support the early diagnosis of neurodegenerative diseases.
🛠️ Technical Skills
I have hands-on experience with:
- Python, C, C++, C#, SQL, JavaScript
- PyTorch, HuggingFace Transformers, TensorFlow
- Git/GitHub, Docker, Unity (ML-Agents), Streamlit
- Google Cloud, SkyPilot, Weights & Biases
Academic Background
- 🎓 M.Sc. in Artificial Intelligence — UFMG (ongoing)
- 🎓 B.Sc. in Information Systems — UFMG
- 🎓 B.Eng. in Control and Automation Engineering — PUC Minas