Data Scientist building practical, evaluation-driven machine learning systems.
I use Python, statistical analysis, and machine learning to turn complex datasets into reliable models and actionable findings, with a focus on NLP, optimization, and reproducible evaluation.
Data Science & Machine Learning Projects
Phishing Email Detection Model
Built an end-to-end LSTM text-classification pipeline for detecting phishing emails. The workflow includes text cleaning, tokenization, padded sequence handling, stratified data splitting, model training, and held-out evaluation across approximately 18,000 emails.
- End-to-end NLP pipeline development
- PyTorch model implementation
- Reproducible train/validation/test evaluation
- Classification-threshold and error analysis
Catan Multi-Agent RL
Built a rules-focused four-player Catan simulation using the PettingZoo AEC interface. The current implementation provides the environment, legal-action handling, turn sequencing, and game-state foundation required for future multi-agent self-play experiments.
- Multi-agent environment design
- PettingZoo AEC interface
- Game-state management & legal-action validation
- Turn sequencing & foundation for self-play RL
Delivery Route Optimization
Developed an optimization-based approach for evaluating delivery assignments and routes under operational constraints. The project focuses on translating a real-world logistics problem into measurable objectives, constraints, and decision rules.
- Baseline approach: TBD
- Selected algorithm: TBD
- Validation method: TBD
Core Capabilities
Machine Learning
- Regression & Classification
- Clustering & Ensemble methods
- Model comparison
- Cross-validation
- Hyperparameter tuning
- Evaluation and error analysis
NLP & Deep Learning
- PyTorch
- TensorFlow/Keras
- LSTM networks
- Text classification
- Embeddings & Tokenization
- Sequence preprocessing
Data Analysis & Engineering
- Python (pandas, NumPy)
- SQL
- Data cleaning
- Feature engineering
- Missing-value handling
- Reproducible pipelines
Visualization & Communication
- Matplotlib & Seaborn
- Tableau
- Technical reporting
- Stakeholder-focused metrics
- Clear presentation of model results
Education & Technical Background
M.S. in Data Analytics
Development Tools
A concise overview of the technical environment I use to build reproducible projects:
About Andres
I am a data scientist with an M.S. in Data Analytics and a background working in enterprise technology environments. I build practical machine-learning solutions with an emphasis on reproducible evaluation, readable code, transparent metrics, and results that can be translated into business decisions. My current work includes NLP classification, optimization, and multi-agent simulation.
Get In Touch
Feel free to reach out to discuss machine learning, data roles, or project collaborations.