Brian González Beltrán

Data & AI Engineer · Bogotá, Colombia

About

Industrial Engineer from Universidad de los Andes with 5+ years building data systems across banking, fintech, and the public sector. Currently transitioning into AI Engineering — building production-grade LLM applications, RAG systems, and automation pipelines on my own infrastructure.

This site hosts live projects that I designed, built, and deployed end-to-end. Everything here runs on a self-managed Oracle Cloud server using Docker, FastAPI, Ollama, and ChromaDB.

Live Projects

RAG Question-Answering API

A retrieval-augmented generation system that answers questions based on real documents instead of hallucinating. Embeds documents into a vector database, retrieves relevant context, and grounds LLM responses in source material.

Try the API → View code →
FastAPI · ChromaDB · Ollama · phi3:mini · nomic-embed-text

LLM Chat API

A FastAPI service that serves a locally-hosted language model through a simple REST endpoint. Demonstrates LLM integration, prompt handling, and async request management.

Try the API → View code →
FastAPI · Docker · Ollama · phi3:mini

API Health Monitoring System

Production monitoring system built for the Bogotá Secretariat of Economic Development. Detects external API failures (Google Geocoding, Registraduría) and sends automated email alerts via webhooks before users notice the problem.

Learn more →
Python · n8n · Webhooks · YAML configs

Stack

Languages: Python, SQL (advanced), R, VBA

AI & Data: FastAPI, ChromaDB, Ollama, MS Fabric, Power BI, Snowflake, RAG architectures

Infrastructure: Docker, nginx, Linux, Oracle Cloud, n8n, Git