- Leading a 20-person AI and data team.
- Driving major delivery programs, including company-wide AI tooling adoption.
- Owning the intersection of AI strategy, platform capability, and execution in a regulated environment.
About
Operator first, technologist second.
The site used to read like a generic AI portfolio. That was the wrong frame. The real story is leadership: building teams, shaping platforms, and making AI useful at company scale.
I currently serve as Director of AI and Data at Renta 4 in Madrid, where I lead a 20-person team and drive major initiatives around AI adoption and data capability.
Before that, I worked across startups and scale-ups in Madrid, London, and Abu Dhabi, building production systems in NLP, MLOps, backend platforms, and data infrastructure. The pattern across those roles has been consistent: take ambiguous technical problems, put an operating model around them, and ship something that lasts.
My background spans hands-on machine learning engineering, API and platform development, cloud infrastructure, and technical leadership. That mix matters because most AI problems fail at the boundary between model, product, and execution.
- Leadership Org design, hiring, mentoring, delivery standards, and cross-functional execution.
- AI systems LLMs, agent workflows, RAG, model serving, evaluation, and production reliability.
- Platforms Data pipelines, APIs, MLOps foundations, cloud infrastructure, and internal tooling.
- Business context Translating technical capability into adoption, process change, and measurable operational impact.
Experience
Recent roles and operating scope.
Current role first. Previous roles are drawn from the CV, with the inflated filler stripped out.
- Built production LLM conversational platforms processing 10k+ daily user interactions for B2C financial services applications.
- Developed autonomous agents for document processing and report generation, cutting manual processing time by 75%.
- Implemented RAG architectures across ElasticSearch, Neo4J, and MongoDB, and improved response time from 40s to 10s.
- Refactored LLM-based data pipelines and integrated Prefect orchestration, reducing ETL processing times by 40% over roughly 50k records per day.
- Hardened infrastructure by moving credential management to AWS Secrets Manager.
- Deployed LLM inference with ONNX optimization and Triton Server, enabling major cost and time improvements.
- Architected a greenfield AI platform using Python, FastAPI, and LLMs, serving 40k+ users with 99.9% uptime.
- Built end-to-end MLOps pipelines with Kubernetes, GCP, and Vertex AI, reducing time to production by 70%.
- Led a five-engineer cross-functional group while partnering with leadership on roadmap and technology strategy.
- Delivered end-to-end ML and NLP systems in production environments.
- Worked across model development, APIs, deployment, and engineering standards.
- Established internal capability around ML engineering practices and Python standards.
Impact
What the CV actually proves.
These are the concrete signals worth keeping visible. They say more than any generic “featured projects” section.
20 people
Current leadership scope at Renta 4 across AI and data.
10k+
Daily user interactions supported by production LLM systems in a prior financial services context.
75%
Manual processing reduction through autonomous agent workflows.
70%
Time-to-production improvement delivered through MLOps platform work.
Contact
Direct and simple.
Available for leadership conversations, advisory work, and serious AI platform discussions.
Based in Madrid. Focused on AI leadership, data capability, production systems, and enterprise adoption.
The current homepage now reflects your actual profile. The PDF CV itself still appears to be outdated and anchored to the Abu Dhabi role, so that document should be refreshed next.