AI Infrastructure Engineering in Portugal · Miguel Enes

AI infrastructure work connects model integrations to dependable delivery: repeatable pipelines, tested deployments, and operational knowledge that remains with the client team. From Porto, I do this through luminous-giant.com and draw on a delivery record that includes modernizing ML infrastructure for a platform team.

I am Miguel Enes, founder of luminous-giant.com, with engineering leadership experience at Lazada, NEW WORK SE, and MIR Insight. I apply that experience to AI and platform infrastructure for Portuguese and European clients.

What AI infrastructure delivery covers

A working AI system is mostly plumbing done well:

• Containerized workloads. ML pipelines packaged in Docker so they scale horizontally instead of living on one machine.
• CI/CD pipelines. Automated build, test, and deployment with GitHub Actions, so shipping is routine rather than an event.
• LLM integrations. Python services that connect large language models to your data for extraction and task automation — wired into your stack, not a sandbox.
• Operational ownership. Monitoring, deployment processes, and documentation that let your team run the system without me on call.

Evidence from the delivery record

These are delivered outcomes from my own record, each scoped to its context:

• At MIR Insight (Porto, 2025), I led the platform team through ML infrastructure modernization: containerized ML pipelines for horizontal scaling and reduced deployment time by 40% through Docker and GitHub Actions automation. The same historical rebuild reduced production incidents by 35% through automated testing and code reviews.
• At Bamilo (2016–2017), as CTO of one of Iran's largest marketplaces, I achieved horizontal scalability and automated deployment for peak traffic, and cut cloud infrastructure costs by 30% by replacing proprietary tools with open-source alternatives.
• At luminous-giant.com, I build and maintain backend services for Portuguese clients using Laravel and Rust, manage infrastructure and deployment for a network of media sites, and ship Python LLM integrations that automate data extraction.

The venture details are in the luminous-giant.com dossier, and the delivery model is described in how luminous-giant.com ships client-owned systems.

Who this is for

• Portuguese and EU companies piloting AI that now needs pipelines, monitoring, and cost control — not another prototype.
• Media and content operators with publishing workflows that should run themselves: luminous-giant.com already automates these pipelines for a network of media sites.
• Platform and data teams whose ML workloads need containerization and CI/CD discipline from someone who has done the rebuild before.

If you need the systems-ownership argument spelled out, continue with client-owned AI systems.

Work with me

Infrastructure work starts with an audit of what you run today: pipelines, deployment path, and where manual effort hides. Then we containerize, automate, and hand over something your team owns.

• Visit luminous-giant.com for the consultancy.
• Or discuss a build with Miguel directly.

For the engineering patterns behind these systems, read production AI workflows: a field guide.