About

I lead AI at Shakers, a flexible-talent marketplace in Madrid with 450+ enterprise clients and 15,000+ verified freelancers. I own the technical architecture and product direction of our AI surface — Matchmaking, Talent Insight, and the ShakersAI platform underneath. My job is to ship AI that recruiters trust, candidates can challenge, and regulators can read.

That framing isn't marketing. Shakers' Matchmaking system is listed in the OECD.AI Observatory catalog as a reference matchmaking system. We are 1 of 12 Spanish companies selected by Spain's Ministry of Digital Transformation (DG IA) for the EU AI Act regulatory sandbox, where I run the 8-month adaptation programme in direct collaboration with BBVA, DG IA, and AESIA. Shakers was in the first cohort to obtain the Adigital Algorithmic Transparency Certificate. I led that, too.

Before all of this, I trained as an applied statistician — psychometrics, multivariate methods, measurement theory. PhD work with 2,780+ citations and an h-index of 19. That foundation is the lens I bring to AI now — most of what's called “LLM evaluation” is a statistics problem with new vocabulary, and I think more operators should treat it that way.

What I do at Shakers

• Matchmaking — LLM-based talent-project ranker with explainable, auditable reasoning. Powers 7,500+ projects.
• Talent Insight (Alma) — conversational interview system that turns CVs into interactive, context-aware profiles. In production.
• ShakersAI — the AI platform under everything above. Built from zero. Self-hosted Qwen on AWS via vLLM for privacy-sensitive flows; OpenAI, Gemini, Anthropic for the rest; Langfuse, Datadog, and Promptfoo for observability and eval.
• AI Act readiness — risk classification, quality management system, technical documentation, all the unglamorous paperwork that turns a model into a defensible product.

Outside the role

I speak publicly when the topic is one I can speak about honestly — most recently AI Summit Barcelona, ProductTank Madrid, and the AMETIC IA awards. The HR Review interview is a decent snapshot of how I think about deploying AI without stripping out the human part.

This site is the long-form half of what I do. If you want a faster path in, the book-a-chat link in the footer works.