I'm a Full-Stack and Machine Learning Engineer based in Brazil, with 5+ years of experience working across education, government, and academia. I hold a Bachelor's in Computer Science from UFJF, with a thesis on ML applied to reading fluency assessment, plus a postgraduate certificate in databases.
My core stack spans Python, Java, TypeScript/JavaScript, React, Spring Boot, Node.js, PostgreSQL, and Docker — giving me strong end-to-end coverage from frontend to backend.
On the ML side, I have hands-on production experience with PyTorch, Wav2vec2, Kaldi, and Scikit-learn, primarily applied to NLP and automatic speech recognition. I trained and deployed ASR models that reduced manual processing time by 50%, and built a subtitle generation proof-of-concept using transformer-based models — work that sits directly on the same architecture shared by modern LLMs.
As a researcher, I'm a published first author and Best Paper Award winner at SBIE 2022, with three co-authored papers on ML applied to children's reading fluency. That experience taught me to evaluate technical work with academic rigor — something I bring into every engineering decision I make.
As a leader, I currently serve as Tech Lead at BrainyIT, coordinating a team of 8 engineers building Eduplay — a national educational video platform deployed across all 5,570 municipalities in Brazil, in partnership with RNP, the Brazilian government's research and education network.
What sets me apart is the combination of research rigor, production ML experience, full-stack depth, and team leadership — all applied to real-world problems at national scale. I'm also actively expanding into LLM fine-tuning, RAG pipelines, and the Hugging Face ecosystem, building naturally on my existing transformer-based foundation.