engineer, founder,
product leader
Everything I've built comes down to one problem: getting people to trust technology when the stakes are high. I studied Aerospace Engineering at UPV (opens in new tab) because I love complex systems — rockets, satellites, machines where failure isn't an option. The parts interested me less than the whole: the team, the strategy, the call on what to build. That pushed me toward product, and then toward starting a company.
In 2017 I co-founded Stratolloon (opens in new tab) with Alex Caballero (opens in new tab). We started in hardware: a system that landed stratospheric balloons on a precise spot using motors and an autopilot instead of a parachute. Then we pivoted to software — a satellite-data analytics platform for insurance and geointelligence. We won the Galileo Masters — presented by the European Commission and ESA — in 2017 and 2018, and earned recognition from Banco Santander and Indra, among others.
Awards are good credentials. They don't build a business. After years of building and iterating, we made the call to stop — knowing when to stop is a product decision too.
I spent the next three years at Exponentia (opens in new tab) as Product Manager, shaping data and product strategy across luxury retail, agri-food, and logistics. My biggest project there was Datocles, an AI recommendation engine for cross-sell and forecasting. In its first year it drove €20M+ in client cross-sell revenue traced to its recommendations, and won a Forbes Spain innovation award in 2022. On the side I mentored early-stage founders and helped build an entrepreneurship program at UPV (opens in new tab) with UC Berkeley, evaluating 200+ applications and owning the relationship between the two universities.
Today, as Head of Product & AI at GPTadvisor (opens in new tab), I'm back on that same challenge. I set the roadmap and lead three teams — product, data, and customer success. That's 25 people and three team leads, and I report to the CEO. We build generative AI for private banking and wealth management, about as compliance-bound as software gets. Major banks across Europe and Latin America run it, and ~9,000 advisors use it every day.
Most of the work behind that trust is keeping the LLM under control: evaluation that compliance will sign off on, vendor due diligence, security review, DORA, the EU AI Act. And a deterministic engine that keeps the model out of every portfolio calculation, so no figure an advisor sees is model-generated. That's the difference between a pilot and something a bank puts in front of a client.
- banks across Europe & LATAM
- 15+
- advisor satisfaction
- 97%
- time saved in portfolio analysis
- ~85%
When I'm not at GPTadvisor, I build the tools I wish existed: Apophis Tracker (opens in new tab), sancho.vc tools (opens in new tab).