Ferran Vidal-Codina
I build production AI, data systems and models for complex, high-stakes domains.
My career spans nine years across research and engineering, including production work for FIFA, Google, the English Premier League, the Golden State Warriors, FC Internazionale, and US Soccer. I turn difficult data and modeling problems into systems people can trust and use.
Selected projects
Problem → approach → measurable resultGCP · RabbitMQ
FIFA World Cup 2022 02 Extracting match events from tracking data July 2019 – June 2023A provider-agnostic system that automatically detected and extracted structured events from player and ball tracking data. 450+ matches
90%+ detection rate
Deployed at the 2022 World Cup 03 Digitizing NFL playbooks 2025 – 2026Synthetic-data generation, semantic segmentation, and route tracing for football playcards. Zero real training labels
Multi-GPU Tensorflow
>80% real-world IoU 04 Conversational sports analytics2025 – presentNatural-language analysis grounded in accurate structured data for NBA, MLB, and NCAA basketball. Prefect · Firebase
BigQuery · DuckDB
RAG · LangChain · OpenAI
Across the stack
I’m happiest when I can follow a problem end to end: get the data from wherever it lives, validate it and make sure it’s consistent, build the model or workflow, and extract insights and value for the stakeholders who need them. I’ve done the work at every layer of the data value chain, so I know exactly where handoffs tend to break down.
How I work
I start with a thorough understanding of the problem and its requirements, then build small and put the work in front of the people who need it early. From there I stay close to the rough edges: the data that exhibits inconsistencies, the cases a model misses, and the general messiness that accompanies real systems. That is where attention to detail matters most, and where I spend real time translating what the data means for stakeholders who may not be technical.
Recognition
Paper of the Year
International Sports Engineering Association and Springer Sports Engineering, for Automatic event detection in football using tracking data.
La Caixa Foundation Fellowship
$150,000 fellowship supporting graduate study at MIT.
Contact
Based in Hawai‘i and working remotely across US time zones. Reach me at vc.ferran@gmail.com or on LinkedIn. Research record: ORCID.