Data Scientist with 5+ years delivering end-to-end solutions for retail and adjacent industries. I cover the full stack—from data engineering/ETL to modeling (ML/DL/NLP), validation, and production/MLOps—primarily in Python and SQL. Strong track record building scalable pipelines (BigQuery, Databricks, Spark) and running cloud-native deployments (Google Cloud, Azure, AWS), tying technical metrics to business outcomes.

🔎 NLP at scale: customer review mining, sentiment/topic classification, and LLM workflows (prompting/embeddings) for actionable insights.

☁️ Cloud & Production: data orchestration, experiment tracking, and model deployment on Vertex AI and Azure ML.

📈 Impact-driven analytics: predictive modeling (churn, recommendation, time series) and executive dashboards (Power BI, Qlik Sense, Looker Studio) for decision-making.

One-liner (optional): Data Scientist (5+ yrs) — NLP, forecasting & recommenders in production; Python/SQL, BigQuery/Databricks/Spark, Vertex AI/Azure ML; business-impact focus.

Programming Languages

• Python, R, SQL

Data & ETL

• Pandas, NumPy, PySpark, Spark, Hadoop
• SQL Server, BigQuery, Databricks

Statistics & Mathematics

• Descriptive & inferential stats, probability, regression, hypothesis testing

Machine Learning

• Regression, Classification, Clustering (K-Means, DBSCAN, Hierarchical)
• Ensemble methods: Random Forest, XGBoost, LightGBM, CatBoost
• Feature engineering & selection, dimensionality reduction (PCA, t_SNE)
• Recommenders, anomaly detection, hyperparameter optimization (Optuna, Bayesian)

Time Series

• ARIMA/SARIMA, Prophet
• Deep Learning: LSTM, GRU

Deep Learning

• MLPs, CNNs, RNNs
• GANs (DCGAN, Pix2Pix, CycleGAN, DeOldify)
• Transfer Learning (VGG, ResNet, MobileNet, EfficientNet)

NLP & Transformers

• Preprocessing (tokenization, lemmatization, n-grams, embeddings)
• BERT (sequence classification), BART (summarization/generation)
• ChemBERTa (biomedical/healthcare text)
• Sentiment analysis, classification, summarization, NER

LLMs (Large Language Models)

• Fine-tuning, Prompt Engineering, RAG
• Embedding-based classifiers, LangChain, LlamaIndex, Gemini, LLaMA

Cloud & MLOps

• Google Cloud (BigQuery, Vertex AI), Azure ML, AWS SageMaker
• Model deployment, CI/CD, monitoring, pipelines (Airflow, Prefect, Kubeflow)

BI & Visualization

• Power BI, Qlik Sense, Looker Studio, Excel
• Matplotlib, Seaborn, Plotly

• CNN Car Damage Severity (MobileNetV2 + YOLOv8) – severity classification and damage detection.
  Repo: https://github.com/RafaelGallo/CNN_Car_Damage_Severity_MobileNetV2_YoloV8
  Deep Learning · CV · Transfer Learning · Detection & Classification

• Bank Customer Churn – ML Pipeline – churn prediction with feature engineering & validation.
  Repo: https://github.com/RafaelGallo/ML_Bank-Customer-Churn
  Classical ML · Classification · sklearn Pipeline

• Auto Insurance Claims – Perceptron/MLP – supervised modeling for claims/propensity.
  Repo: https://github.com/RafaelGallo/Auto-Insurance-Claims-Neural-Networks-Perceptron-MLP
  Neural Nets · MLP · Evaluation

• DeOldify – GAN Colorization – colorizing B/W images with DeOldify.
  Repo: https://github.com/RafaelGallo/GAN_Generative-Adversarial-Network_DeOldify
  GAN · Computer Vision · Inference

• Plant Disease Recognition – CNN (MobileNetV2/VGG16) – leaf disease classification.
  Repo: https://github.com/RafaelGallo/Plant-disease-recognition---CNN-MobileNetV2-VGG16
  Deep Learning · Transfer Learning · Images

• Breast Tumor Cell Nuclei – CNN Segmentation – nuclei/cell segmentation in pathology.
  Repo: https://github.com/RafaelGallo/Breast-Tumor-Cell-Nuclei-Convolutional-Neural-Network-Segmentation
  Segmentation · CNN · Healthcare

• Skin Cancer – U-Net VGG16 – skin lesion segmentation with U-Net (VGG16 encoder).
  Repo: https://github.com/RafaelGallo/Convolutional-Neural-Network-Skin-Cancer-U-Net-VGG16
  U-Net · Transfer Learning · Medical Imaging

• LLM Prompt Engineering – Llama 3 (Sentiment/Climate) – prompt design and sentiment analysis.
  Repo: https://github.com/RafaelGallo/LLM_Engineering_prompt_LLama3_Sentiment_analysis_climate
  LLM · Prompting · NLP

  📚 Kaggle Competitions

• Kaggle Projects
  Works hosted in Kaggle_comp: Notebooks and production pipelines for datasets, including EDA, feature engineering, modeling, and submission for real-world challenges.
  Python · scikit-learn · Ensemble models · EDA · visualization · documented kernels

• FIAP Cognitive Environments
  Hands-on work with cognitive environments, agents, computer vision and NLP.
  Python · CV · NLP · Heuristics
  Repo: https://github.com/RafaelGallo/FIAP_Cognitive_Environments

• Deep Learning Rede Neural FIAP MBA
  Deep Learning models (CNNs, LSTM/GRU), tuning, callbacks, metrics and best practices.
  TensorFlow/Keras · PyTorch · Regularization · EDA
  Repo: https://github.com/RafaelGallo/DeepLearning_Rede_Neural_FIAP_MBA

• Bot Trader LLM Qlearning and RNN
  Trading agent with LLM (prompting) + Q-Learning + RNN, plus backtesting.
  Reinforcement Learning · Time Series · Finance · Backtest
  Repo: https://github.com/RafaelGallo/robo_trader_LLM_Qlearning_RNN

• Portiforio FIAP MBA Data Science
  MBA portfolio (EDA, classical ML, visualization and reports).
  Pandas · scikit-learn · BI · Storytelling
  Repo: https://github.com/RafaelGallo/portiforio_FIAP_MBA_DataScience

• Practical Projects – Data Science Academy
  A collection of mini-projects from the Data Science Academy training: big data with R and Azure ML, Python/Spark, Machine Learning, Business Analytics, visualization, and data engineering (Hadoop/Spark).
  Tasks include churn analysis, recommender systems, fraud detection, sentiment analysis, time series forecasting, and dashboards.
  Repo: https://github.com/RafaelGallo/Projetos_dsa

• Deep Learning Specialization (DeepLearning.AI – Coursera)
  Hands-on implementations across five modules, covering everything from basic neural networks to CNNs, Seq2Seq, LSTMs/GRUs, and optimization strategies.
  Python · TensorFlow/Keras · Backpropagation · CNN · RNN · Regularization
  Repo: https://github.com/RafaelGallo/Deep-Learning-Specailization-DeepLearningAI-Coursera

• LinkedIn: https://www.linkedin.com/in/SEU_USUARIO
• E-mail: [email protected]
• GitHub: https://github.com/RafaelGallo