About

I'm an AI systems engineer and builder from Peru. I build AI products, research experiments, and production software, and I care about understanding how these systems actually behave.
I started coding professionally in 2020. Worked at FITCO (TechStars-backed fitness SaaS), then Roma (education platform), then Deepview where I re-architected their backend to process 4B+ images per year using Kubernetes and real-time ML inference.
My best-known project is Symbiotic AI (680+ GitHub stars) — a personal AI system for memory, reflection, and behavior change. I also built So Long Sucker, AI deception research based on John Nash's betrayal game, which hit the HN frontpage and was picked up by Gigazine Japan.
My range goes from embedded C++ on ESP32 microcontrollers to cloud APIs to full-stack web apps. I learn fast and apply what I learn immediately.
Long term, I'm building toward human-AI symbiosis: systems that help people think more clearly and operate with more leverage. I'm especially interested in agent architectures, AI behavior, and useful systems that matter in the real world.
When I'm not coding, I go for walks to think and listen to music to reset.
Career
Senior Backend Engineer, Deepview
Microservices, Kubernetes, ML inference at scale
Full Stack Engineer, Roma
Education platform
Full Stack Engineer, FITCO
TechStars-backed fitness SaaS