Industrial Engineer from Universidad de los Andes with 5+ years building data systems across banking, fintech, and the public sector. Currently focused on AI Engineering — shipping production LLM applications, analytical pipelines, and agentic systems on self-managed infrastructure.
Everything on this site is live, built end-to-end, and running on a single Oracle Cloud ARM server.
End-to-end analytics pipeline over open government data. Ingests 76K+ business records from ArcGIS REST API, labor market indicators, and quarterly GDP — transforms into a star schema in PostgreSQL, serves 14 analytical endpoints via FastAPI, and renders an interactive dashboard.
Retrieval-augmented generation system that grounds LLM responses in real documents. Embeds and indexes documents into a vector store, retrieves relevant context at query time, and generates answers with source attribution.
FastAPI service wrapping a locally-hosted language model through a REST endpoint. Handles prompt engineering, async requests, and timeout management for production-grade LLM serving.
Production alerting system for the Bogotá Secretariat of Economic Development. Detects external API failures in geocoding and identity services, sends automated email alerts via webhooks before users notice.
Python, SQL, PySpark, R, VBA
FastAPI, PostgreSQL, ChromaDB, Ollama, Snowflake, Power BI, RAG, MS Fabric
Docker, nginx, Linux, Oracle Cloud, n8n, Git, Databricks