Scientist & Entrepreneur.
Valencia, Spain
I'm a biomechanics researcher turned builder. My PhD at Imperial College London studied how people with lower-limb amputations walk.
In fitness tech, I built AI models that recognize and correct exercise form using just a camera, then moved into a role running several product pipelines and automating processes — including one workflow I cut from two weeks down to two hours. After that, I founded my own company: raised funding, built a product, got our first clients, and learned how to run a business.
Now based in Valencia, looking at what's next. I'm drawn to work where science and AI meet — health tech, medtech, biotech, wearables. Happy to talk to anyone working on interesting problems in this space.
Biomechanical outcomes of through-knee amputation, studying implications for surgical decision-making, prosthetic socket design, and long-term health.
Poster presentation at the International Society of Biomechanics Congress 2021: “Ground reaction forces during walking of people with traumatic bilateral major lower limb amputations.”
Also tutored medical students on musculoskeletal conditions, bone and joint disorders, and spinal anatomy.
Founded an orchestration layer for building AI agents and automating business processes, shipped as the product Flero. Raised a pre-seed round, built the product, and landed our first clients — automating hundreds of hours of manual work a month for them.
Developed and executed an innovation strategy alongside machine learning engineers and product teams. Led efforts to streamline internal workflows and shipped prototypes and proofs-of-concept from ideation through implementation.
Developed, calibrated, and refined exercises for the Mvmnt audio fitness app. Contributed to ML models for markerless exercise recognition and segmentation using MediaPipe, and refined calibration based on user feedback.
Sourced academic speakers for Researcher Live events reaching 2.3 million users, and helped create and promote live series events.
An orchestration layer for building AI agents and automating processes with flexibility and cost control — built at AICE Technology. Used by clients to automate lead intake, CRM data entry, and sales collateral generation, saving teams hundreds of hours a month.
A personalised audio fitness app. Built and calibrated markerless, camera-based exercise recognition so the app could adapt to users of all abilities and body types.