Getulio Cesar De Leon Fernandez

I'm an economist in the National Accounts department of the Central Bank of the Dominican Republic, where I help compile GDP under the SNA 2008 framework - reconciling supply and use tables built from the financial statements of thousands of firms and government institutions. Most of that work used to mean six analysts fighting over the same shared spreadsheets. I don't like that kind of bottleneck, so I build the tools that remove it.

That instinct is what led me to a Python application (pandas + xlwings) that now orchestrates the compilation of national accounts by economic activity, freeing roughly four hours of team time a day; a data pipeline that consolidated scattered files into a single table of over 1.3 million records, with a validation script that caught 7,000 year-over-year catalog changes an audit would otherwise have missed; and a disaggregated Supply and Use Table of more than 480 columns with its own optimization algorithm for balancing supply against use.

I'm currently completing a Master's in Big Data Analysis and Visualization (UNIR), and I use personal projects to go deeper into the parts of the stack my day job doesn't require yet - self-hosted infrastructure, LLM-driven automation, dashboards built from public data. The projects below are representative examples, all built with synthetic or public data.