🏛️ Data Analyst II at the University of Miami Miller School of Medicine
📜 BA History → MA International Administration → MS Business Analytics
💬 English and Spanish (native), Catalan (intermediate)
Below is a selection of tools and technologies featured across my repositories and professional experience:
| Category | Tools |
|---|---|
| ⌨️ Languages | Bash C# CSS3 Go HTML5 JavaScript Python R SQL VBA |
| 🛢️ Databases | BigQuery Databricks Firebird MariaDB MySQL Oracle PostgreSQL SQL Server SQLite |
| 🧮 Libraries | BeautifulSoup Chart.js Matplotlib NLTK numpy pandas pdfplumber Pillow Playwright Plotly Puppeteer python-docx requests scikit-learn SciPy seaborn Selenium spaCy statsmodels TensorFlow Tidyverse Tkinter wordcloud XGBoost |
| 📉 BI & Reporting | DAX IBM Cognos Power BI Tableau |
| ♻️ Reproducibility | JSON Jupyter LaTeX Markdown Marimo Mermaid Quarto R Markdown |
| 🤖 AI & LLMs | ChatGPT Claude Gemini Groq Lumo Mistral Perplexity |
| 🏭 Enterprise Systems | Power Automate Qualtrics SharePoint Smartsheet Workday |
| 🔨 Dev Tools | Chrome DAX Studio Docker Git GitHub Actions Hugo Prettier Tabular Editor 2 Tailwind CSS |
| Analysis | What It Does | Why It's Useful |
|---|---|---|
| 📐 Multi-Campus NCLEX-RN Outcomes | 7,635 exam attempts across 19 campuses in a normalized SQLite database, with pass rates and 95% confidence intervals computed in pure SQL and a logistic model fit in R | Shows the whole path from raw data to a defended conclusion, published reproducibly through Datasette Lite so anyone can rerun it |
Outcomes are synthetic over a real structural skeleton. The write-up is explicit about what it can and cannot claim.
| Project | What It Does | Why It's Useful |
|---|---|---|
| 🔎 sql-x-ray | Dumps any SQL database's structure, never its data, as priming context for an LLM | One dependency-free query across 8 engines, structure-only by design, so schema metadata stays shareable |
| 📤 pbi-model-export | Serializes a complete Power BI model to portable JSON: measures, relationships, RLS roles, partition queries | Turns hours of manual model documentation into one script run |
| 🧹 qualtrics-processing-pipeline | Turns a raw Qualtrics Excel export into a cleaned dataset, a codebook, and a documented audit trail | Strips test responses and fixes data types automatically, so analysis starts from something you can defend |
| 🪟 tabs-2-json · Chrome Web Store | Reads the text and metadata of the browser tabs you select and exports them as structured JSON | Turns a pile of open tabs into clean, LLM-ready JSON in one click, without the nav and footer clutter |
| 📖 repo-omnibus | Collects every public repository on a GitHub account into one Markdown file, filtered to what a person would actually read | Hand a whole body of work to a reader or a model at once |