Data Analyst · Automation & Data Pipelines
I design and build data pipelines that automate collection, processing, and analysis — not just dashboards, but reproducible systems that generate reliable data over time.
Python SQL Data Automation CI/CD ETL
"Data is not just analyzed. It must be generated, structured and sustained."
- Automated data pipelines (scraping + storage + CI/CD)
- Data analysis with Python & SQL
- Transitioning towards Data Engineering
- Collection — Playwright, yt-dlp, Selenium, REST APIs
- Processing — Python (Pandas, NumPy), SQL (CTEs, window functions)
- Orchestration — GitHub Actions
- Visualization — Power BI, Streamlit
A comprehensive electoral intelligence system for the 2025 Chilean presidential elections (covering both the first and second rounds across 346 communes within 16 regions). The workflow spans the entire data pipeline: from retrieving official data from SERVEL using custom Selenium-based scrapers, through an analytical layer in SQL Server (incorporating 26 T-SQL queries with CTEs) and bilingual Jupyter notebooks—focused on territorial mapping and analysis (including K-Means algorithms)—to an interactive Streamlit dashboard (available in both Spanish and English) featuring a complex, six-level navigation system. The system includes name normalization across disparate datasets, geographic transition mapping, and strategic analysis of the electoral behavior of voters who supported Parisi, as well as trends from previous elections.
- github.com/adroguetth/Chilean-Presidential-Elections-2025-Analysis — Repository (completed)
- chilean-presidential-election-2025.streamlit.app/ — Online Dashboard
End-to-end ETL pipeline that downloads, enriches, and analyzes YouTube weekly charts. Features multi-source artist enrichment (MusicBrainz, Wikipedia, Wikidata, DeepSeek), 3-layer metadata retrieval (API / Selenium / yt-dlp), and automated bilingual Jupyter notebooks with AI insights. Fully orchestrated on GitHub Actions with 4 integrated workflows.
github.com/adroguetth/Music-Charts-Intelligence — In progress
End-to-end retail sales simulation pipeline that generates realistic synthetic sales data for 5 stores and 25 products, enriched with dynamic e‑commerce events (CyberDay, CyberMonday, Black Friday), multi‑factor demand modeling (seasonality, holidays, promotions, stockouts), and automated daily fact table generation. Features intelligent event date fetching via multi‑source web scraping (12 sources) with DeepSeek arbitration, static master data management, and fully automated daily ETL on GitHub Actions with 3 integrated workflow
github.com/adroguetth/Sales-Monitoring-Engine — In progress
- Docker (containerization)
- PostgreSQL
- Pipeline orchestration (Airflow / Prefect)
- R — exploratory & research-focused analysis
Open to collaborations, freelance work, or just talking data.