Daniel Ochoa

I built six backend systems of my own using Go and Python, with public source code on GitHub. My work includes production APIs and LLM integration with RAG, MCP, and multi-provider orchestration.

For my RAG Document API, I developed multi-provider document search with locally run HuggingFace embeddings, keeping data in-house and avoiding per-token costs. It supports asynchronous ingestion of several document formats and includes a dashboard for uploads and chat.

My Engine API Workflow uses Redis queues for asynchronous jobs with retry policies, and secures endpoints with JWT, RBAC, refresh tokens, and per-user rate limiting. I organized it with a clean layered architecture and added audit logging, metrics, and health checks.

Before focusing on backend development, I served in the Argentine Navy from 2011 to 2025, most recently as Electronic Repairs Supervisor for navigation radar systems. I led a technical team maintaining mission-critical equipment, reduced downtime through fault diagnosis and preventive maintenance, and documented procedures and interventions.