FREME Project

FREME was a European Commission-funded research project aimed at closing a persistent gap in the language technology field: powerful multilingual, semantic technologies existed in research labs, but industry partners had no practical way to access or apply them. I served as tech lead, directing an international team of 15 AI engineers and researchers tasked with bridging that gap.

The Challenge

Research-grade semantic and multilingual technology is often built to prove a concept, not to be adopted by a company with its own systems, deadlines, and non-research priorities. FREME’s mandate was to make that technology genuinely usable by industry partners — which meant the project lived at the intersection of cutting-edge NLP research and the much less glamorous work of packaging, documentation, and integration that determines whether research actually gets used.

My Approach

I led a 15-person international team of AI engineers and researchers, directing the technical work of transferring multilingual, semantic technologies into forms industry partners could actually adopt. That meant balancing the project’s research ambitions against the practical needs of the companies we were building for, coordinating a genuinely international team, and keeping delivery on track against EU project milestones and reporting requirements — a different discipline from a purely internal engineering effort.

Outcome

FREME was rated “excellent” by the European Union on completion — the top rating a Horizon 2020-era research project can receive. Beyond the rating, the project left me with a lasting interest in the specific challenge of technology transfer: how to take research that works in a lab and make it something a company outside that lab can pick up and actually use, a thread that runs through most of the applied, industry-facing AI work I’ve done since.

Focus areas: Multilingual & semantic NLP technologies, technology transfer, international team leadership, EU-funded research delivery.