Hello, I'm

Patrick Martins

Sr Data Scientist · AI Engineer · Consultant

Building intelligent systems at the intersection of Machine Learning, Computer Vision, and NLP. I turn complex data into production-ready AI solutions — from research to deployment.

Featured Projects

Selected work in AI, data science, and software engineering — from production SaaS to academic research.

Fishing Data Analysis
Data Science Spatial Analysis GIS

Deep Analysis of Fishing Data

Spatial and temporal analysis of AIS fishing data. Hotspot detection using Getis-Ord Gi*, clustering, and cartographic visualizations for monitoring and decision-making.

Computer Vision for Claims Automation
Computer Vision OCR NLP

CV for Claims Automation

Intelligent document processing using OCR, computer vision, and NLP. Automates header extraction, signature/stamp detection, and textual analysis for a large health insurance provider.

Legal Case Summarizer
NLP Deep Learning T5/BERT

Legal Case Summarizer

Automatic summarization system for Brazilian Federal Supreme Court (STF) rulings using T5 and BERT. Compares machine-generated summaries with human-written references.

Data Mining in Financial Documents
LLM OCR Automation

Data Mining — Financial Docs

AI-powered invoice processing with OCR, NLP, and LLM workflows. Automates data extraction, classification, and report generation from financial documents.

Road Features Detection
Computer Vision Segmentation Satellite

Road Features Detection

AI-based solution to detect, segment, and reconstruct road networks from satellite imagery. Combines segmentation models with geometric analysis for smart-city applications.

AI Livestock Counting
Computer Vision Drones Detection

AI Livestock Counting

Drone-based livestock counting using computer vision. Automates animal detection and counting with high accuracy, reducing operational time, labor, and animal stress.

About

I'm a Senior Data Scientist and Sr Consultant at KPMG Brasil with deep expertise in building end-to-end machine learning systems — from data exploration and model development to production deployment and monitoring.

My work spans Computer Vision (YOLO, object detection, segmentation, satellite imagery), NLP and LLMs (BERT, T5, GPT, RAG, document processing, summarization), and Data Science (ETL, predictive modeling, data quality assessment, exploratory data analysis).

I have experience with Microsoft Azure (DataBricks, Azure Data Lake, Azure ML, Azure Form Recognizer), SQL and NoSQL databases, VectorDB (PineCone, Chroma, Qdrant), and orchestration tools like Airflow, MLflow, and LangGraph.

I also build full-stack AI products — including SaaS platforms like ResumeMatch, Media Handler, and side projects like Shrimp Dev Blog and Promo Finder — using FastAPI, React, Docker, and Cloudflare.

Skills & Technologies

Python PySpark Pandas NumPy scikit-learn PyTorch TensorFlow XGBoost LightGBM OpenCV NLTK Transformers SHAP Plotly Seaborn FastAPI LangChain LangGraph CrewAI LLMs RAG NLP ETL SQL PostgreSQL MongoDB Redis VectorDB Docker Git GitHub Azure Azure DevOps Airflow MLflow Superset

Get in Touch

Open to consulting, collaborations, and interesting projects.