Not a Pokémon 😅, my name is Yan Enrique and I'm a Data Scientist from Brazil...

"Data is the new oil, but intelligence is the new refinery."

For 7 years, my world was defined by pixels. As a freelance Pixel Artist with over 600 completed projects and a 4.8-star rating from over 380 reviews, I learned a fundamental truth: a single misplaced pixel can ruin an entire composition. Today, I apply that exact same obsessive attention to detail to my daily routine as a Data Scientist.

Instead of placing colors on a canvas, my main brush is Python. My day-to-day involves getting my hands dirty with raw, unstructured data—cleaning and transforming it using Pandas and NumPy to ensure every 'pixel' of information is perfectly aligned. I craft visual narratives through Exploratory Data Analysis (EDA) using Matplotlib and Seaborn, and I engineer predictive models with Scikit-learn, XGBoost, and Tensorflow.

Whether I'm meticulously tuning hyperparameters, fine tuning a model, handling an outlier, processing natural language with NLTK/spaCy, or designing an intuitive dashboard, the bridge between my past and present remains the same: I build functional, precise, and visually compelling solutions.

• Programming Languages: Python R SQL JavaScript Dart HTML5 CSS3
• Data Manipulation & Analysis: Pandas NumPy PySpark SciPy Pingouin Statsmodel
• Machine Learning & AI: Scikit-learn XGBoost PyTorch Hugging Face NLTK spaCy Scikit-image
• Data Visualization & BI: Matplotlib Seaborn Tableau Looker Studio Streamlit Excel
• Design & Workflow: Git Figma UX/UI Design Scrum Kanban Data Storytelling

This is where the magic happens. Explore my projects to see how I build bridges between questions and answers.

➡️ Access My Portifolio Page here ⬅️

• 📈 E-commerce Purchase Intent Prediction: End-to-End Data Science Project: From E-commerce Browsing Behavior Analysis to the Deployment of a Conversion-Focused Predictive Model.
• 📈 Retrieval-Augmented Generation for Digital Humanities: A Python and Streamlit-based academic assistant designed to query and answer questions from a database of Digital Humanities articles and citations. Developed to organize my master's degree readings, the system leverages state-of-the-art LLMs for embeddings and text generation, ensuring accurate answers with proper source attribution.
• 📈 Credit Card Clustering & Segmentation: Credit card customer segmentation using Unsupervised Machine Learning (K-Means), optimized via Optuna and interpreted with Explainable AI (SHAP).
• 📈 Credit Card Customer Segmentation API: This API provides endpoints for real-time inference, classifying credit card customers into specific clusters (personas) based on their financial data. Additionally, it integrates with a Streamlit interface, allowing end-users to intuitively test the predictive model live on Hugging Face Spaces.
• 📈 Bank Intent Classifier — Fine-tuning DistilBERT on Banking77: An end-to-end NLP project that classifies customer banking queries into fine-grained intent categories. The pipeline compares a classic machine learning baseline against a fine-tuned DistilBERT transformer, with all experiments systematically tracked using MLflow. For production deployment, the system is fully containerized with Docker, featuring a modular FastAPI backend for model inference and an interactive Streamlit web interface deployed live on Hugging Face Spaces.
• 📈 Korean Bakery Sales: This project showcases a full-year sales analysis of a bakery, based on 10,840 cleaned and structured transactions. The main goal was to turn raw data into clear visual insights, using UX/UI design principles to improve readability.

• 🌐 EconStudy: Interactive web platform for studying Economics subjects. Transforms lecture notes into a searchable knowledge base with 30+ structured concepts.
• 🎮 Candy Mayhem: A turn-based text RPG where candy-themed creatures battle in a layer cake dungeon. Befriend them, fight rivals and explore the delicious universe.
• 📄 UX Briefing: A simple command-line tool built with Python that integrates AI to helps you quickly create a structured project briefing by asking a series of key questions.
• 📝 To-Do List: A command-line To-Do List application in Python that saves tasks to a text file.

Ready to transform your raw data into refined intelligence?

• LinkedIn: Yan Enrique Linkedin
• Email: enrique4work@gmail.com
• Portifolio: Yan Enrique Portifolio

Let's build something memorable together.