Hugo Romero
Back to About

AI Engineer

Orange Madrid, Spain 2023/09 – 2026/03
Context

Joined Orange's AI team when it was still 4 engineers, in an environment with no established patterns for GenAI production systems, technology changing every few weeks, and constant need to engage stakeholders to define scope and justify decisions.

Outcome

Built the technical foundation the team continued to rely on as it grew, delivering MLOps pipelines, LLMOps governance, agentic workflows, and internal platforms in production.

Team Growth
4 → ~30
LLM Providers
4+
Systems in Prod
Multiple

About the Role

I joined Orange’s AI team when it was still 4 engineers, working at the forefront of Generative AI with no established patterns for building these systems in production. The technology was changing every few weeks, and the role required constant engagement with stakeholders to define scope and connect technical decisions to real business outcomes.

I built the technical foundation the team continued to rely on as it grew. That included MLOps pipelines, LLMOps governance, cloud infrastructure patterns, and internal platforms for monitoring and budget control.

Key Achievements

  • Built comprehensive MLOps pipelines on GCP covering the full ML lifecycle from training to monitoring
  • Developed agentic workflows for task automation using MCP servers
  • Created and maintained Terraform modules for cloud-native AI infrastructure
  • Implemented LLMOps governance covering prompt versioning, budget control, and RBAC
  • Administered LLM access across Azure, GCP, OpenAI, and LiteLLM
  • Built internal platforms for monitoring, alerting, and budget automation

Technical Stack

Python, Kubernetes, GCP, Terraform, GitHub Actions, LiteLLM, Azure OpenAI, Pydantic, FastAPI, Docker, Helm, LangGraph, ADK, Vertex AI, MCP, LangChain, ArgoCD, n8n, Airflow.