- M.Sc. in Statistics — Universidad del Valle, Colombia
- Specialized in: survival analysis, predictive modeling, MLOps, and data engineering
- Build end-to-end ML pipelines: from feature engineering and model training to REST APIs and containerized deployment
- Currently working with: Databricks (medallion architecture), FastAPI, Docker, DVC
- Also an instructor — teaching Statistics, Python, and Machine Learning at university level since 2019
- Ask me about: statistical modeling, MLOps pipelines, goodness-of-fit testing, or Python/R packages
Languages
ML & Data Science
MLOps & Infrastructure
Data & Cloud
Apps & Visualization
| Project | Description | Tech |
|---|---|---|
| pdist | Python package to automatically identify the best-fit probability distribution. Implements KS, Anderson-Darling & Chi-square goodness-of-fit tests, AIC/BIC criteria, and visualizations. Supports 9 continuous + 4 discrete distributions. | Python · SciPy · Matplotlib |
| itseries | R package for analyzing irregularly spaced stochastic processes — built during M.Sc. research in Statistics. | R |
| car_predict | End-to-end ML pipeline for car price prediction with model versioning, REST API, and containerized deployment. | Python · DVC · FastAPI · Docker |
| ETL_scraper | Automated ETL pipeline extracting vehicle data, processing with Python, and loading to AWS S3 via CI/CD. | Python · AWS S3 · GitHub Actions |
| cluster-app | Interactive web app for credit card customer segmentation using unsupervised learning, deployed on HuggingFace Spaces. | Python · Gradio · HuggingFace |
| meli_scrapper | Web scraper for all products published on Mercado Libre, containerized and deployed with CI/CD. | Python · Docker · CI/CD |
| pptex | Docker-based toolkit for generating LaTeX presentations and reports on any OS — no LaTeX installation required. Supports pdflatex, xelatex, lualatex, and watch mode. | Docker · LaTeX · Shell |
