Enrique Rubio

What I work on

My day job is Senior Big Data Platform Administration in banking and payments — currently at Redsys, on Cloudera CDH/CDP platforms, with production security, high-availability operations, and L3 support for payment infrastructure. Seven years of that, across banking and payments environments — the full arc is on the homepage.

Alongside that, I build AI agents and AI automation on my own time: systems where an agent (often built on Claude / Claude Code) takes real actions against real infrastructure under explicit human-approval gates, orchestrated with tools like n8n for the workflow/integration layer. More on that approach → or see how I architect the approval gates and audit trails themselves on AI Agent Architecture. The same discipline carries over from the data- platform work above: versioned, observable, safe to fail loudly rather than fail silently.

What I've built

A short, honest list — see Projects for more detail on each:

• An internal AI agent operations system (personal orchestration infrastructure) with a human-in-the-loop review UI for anything an agent does autonomously.
• Traffic OS — a private content-operations application built around explicit human approval at every publishing step.
• GlossAtlas — an independent cross-brand paint finish comparison utility.
• CompatSense — connectivity and compatibility intelligence tooling.

These are operated individually or under Aidtake Labs, the umbrella I build and ship this work under — not a large company, not a funded startup pitch.

How I work

• Technical leadership by doing the work, not just directing it — I design the architecture and write the code for the systems I run.
• I don't automate judgment calls. Publishing, deployment, and anything irreversible keeps a human approval step.
• Fewer systems, built correctly, beats a pile of demos.

Get in touch

For project inquiries, collaboration, or anything else, see the contact page.