Resume

Welcome to my resume! The resume is available as .html and .pdf.

CONTACT

Research focus on computational neuroscience and random graph modeling for neural and social networks.
Coursework: Graph Theory, Statistical Methods, Deep Learning, NLP, Parallel Computing.

MSc/PhD-level coursework: Information Theory, Stochastic Processes, Bayesian Inference, Machine Learning.
Research projects: lattice-based cryptography (CNPQ Scholarship 2021–2022) and information geometry for ML (CNPQ Scholarship 2022–2023).

Designed and shipped monitoring infrastructure that tracks the health of 3 000+ ML features, boosting model reliability.
Built diffusion-based generative models for tabular feature synthesis, improving downstream classification metrics by ~1%.
Pretrained and finetuned LLM-inspired credit assignment models that unlocked R$100M+ in approved client credit.
Led exploratory AI projects spanning self-supervised learning, graph neural networks for fraud, and LLM interpretability.

Helped develop linear optimization software for asset inventory management across global entities.
Engineered reports and analyses that reduced redundant asset movements and improved operational efficiency.

Produced mathematical analyses of derivatives strategies to evaluate opportunities and risk/return trade-offs.

Introduction to Lattice-based Cryptography — investigated reliable lattice constructions for secure communications (CNPQ Scholarship 2021–2022).
Information Geometry for Machine Learning — studied differential geometry formulations of statistical manifolds and ML loss landscapes (CNPQ Scholarship 2022–2023).

Logit Graph: Library for probabilistic graph generation via logit-based edge processes with benchmarks against state-of-the-art statistical models.
Horror Movie Recommendation App: FastAPI service that blends semantic retrieval, IMDb priors, reranking, and integrates UI, PostgreSQL, CI/CD, and payments.
Auction RAG Assistant: LangChain + MongoDB chatbot (Flask UI) that navigates hundreds of auction documents through retrieval-augmented generation.
Graph Similarity via Representation Learning: Contributed ML architectures and baselines for graph distance estimation problems.

Programming: Python, Matlab, C, Java.
Tools & Platforms: Google Cloud Platform, Docker, Git, Linux, Kubernetes, Spark.
ML/AI: PyTorch, TensorFlow, Scikit-learn, Hugging Face, FSDP training.
Data & Visualization: SQL, Matplotlib, Seaborn, Plotly.
Languages: Portuguese (Native), English (Fluent C2), French (Basic A2).