Gabriel Martins
Data Scientist | Machine Learning Engineer | AI Researcher
👋 Hi, I'm Gabriel, based in Rio Branco, Brazil. I leverage Machine Learning and AI to build solutions that deliver real-world value, focusing on Deep Learning (CV & NLP), LLMs, and robust data practices.
🔭 Currently leading research at PAVIC Lab on synthetic data generation for HDR imaging using Diffusion Models & Few-Shot Learning.
🌱 Dedicated to improving AI systems by tackling data scarcity, bias, and quality challenges
💡 Experienced in applying ML/LLMs in legal tech and software testing domains.
🛠️ Tech Stack & Skills
- Languages: Python | SQL | JavaScript/TypeScript
- ML/DL/NLP: PyTorch | TensorFlow | Scikit-learn | Pandas | NumPy | LangChain | Hugging Face | NLTK/spaCy
- Specializations: Computer Vision | NLP | LLMs (RAG, Fine-Tuning) | Diffusion Models | Reinforcement Learning | MLOps
- Tools & Platforms: Azure, AWS (SageMaker, Lambda) | Docker | Git | GitHub Actions | GitLab CI/CD | Vector DBs | Linux
- Web Frameworks: React | Node.js/Express | Angular | Nest.js | Django
📜 Publications & Research Contributions
- Martins, G. et al. "A Review Towards Deep Learning for HDR Reconstruction." (Under Review, Journal of Applied Sciences - MDPI, 2025).
- Martins, G. et al. "A Method for Regression Testing Plan Ordering..." (Preparing for Submission, CIbSE 2025).
