AI & Automation Engineer
Integrations · Chatbots · AI Agents
I build real automation and artificial intelligence solutions — from autonomous agents and data pipelines to production APIs and full-stack applications with LLMs. Bogotá, Colombia 🇨🇴 · Remote.
Tech stack
What I've built
🎯 Problem it solves
Manually analyzing stocks takes hours: opening sources, copying data, writing the analysis. This agent does it end-to-end with zero human intervention.
⚙️ Why these technologies
Python for its scraping and data ecosystem. Gemini API for its ability to reason over complex financial data. HTML for portable reports that require no installation.
Reduces full stock analysis from ~3 hours to under 5 minutes.
🎯 Problem it solves
Searching hundreds of internal documents is slow and imprecise. This system answers questions in natural language by querying only the relevant content.
⚙️ Why these technologies
Chroma DB for vector semantic search (more precise than keyword search). OpenAI for generating contextualized responses. Transcription to feed the knowledge base from audio without manual transcription.
Eliminates manual document searches: instant answers vs. 20–40 min of manual review.
🎯 Problem it solves
Manually processing and organizing notes is repetitive and error-prone. This system detects changes and runs AI workflows automatically without the user having to trigger them.
⚙️ Why these technologies
LangGraph models workflows with conditional logic and persistent state, impossible with simple API calls. The file monitor triggers the workflow only when real changes occur.
Automates organization tasks that previously took 30–60 min of daily manual work.
🎯 Problem it solves
Extracting key information from long PDF documents requires reading full pages. This platform lets you ask direct questions about any PDF and get precise answers instantly.
⚙️ Why these technologies
Next.js for a reactive SSR frontend. Prisma for typed database management. Docker so anyone can deploy without configuration. Webhooks for external service integrations.
Goes from reading a full PDF (~1 hour) to getting the exact answer in seconds.
🎯 Problem it solves
Need for a scalable, documented backend that can connect to any frontend or external service, with persistent data management, production-ready.
⚙️ Why these technologies
FastAPI for its speed and automatic documentation (Swagger). Clean architecture for long-term maintainable code. Docker so deployment is identical in dev and production.
Auto-generated API docs eliminate hours of manual documentation per endpoint.
🎯 Problem it solves
Creating a new Python project with standard structure, linters, tests and Docker takes 1–2 hours. This tool generates everything in seconds with a single command.
⚙️ Why these technologies
Typer for an ergonomic CLI with autocompletion. Rich for visual terminal output. Packaged with Hatchling for global installation as a standalone tool.
Reduces initial project setup from ~2 hours to under 30 seconds.
Let's talk
Have an automation or AI project? I'd love to hear about it. Available for remote work from Bogotá, Colombia.