AI strategy, applied language technologies, and production-grade systems for organizations and learners.
AI, software engineering, and learning systems

Following the creation of Eienn.bzh, my MVP project for a Breton language acquisition platform, I created Fluens.co. The SaaS (Software as a Service) platform was designed to help organizations to create, manage and distribute language learning content effectively and at scale. Key features included: a content management system (CMS), a user-friendly interface for both content creators and learners, analytics and a complex payment integration with Stripe to handle subscriptions and transactions securely. Stack: Vue (+ Radix Vue), Firebase (+ node.js).

Started as my Master's thesis project, Leksis is an adaptive vocabulary learning system that leverages machine learning and deep learning techniques to quickly evaluate language learners's vocabulary level efficiently. Key techniques and algorithms included: a recommendation engine (based on the Elo rating system's logistic model), LSTM cells (PyTorch) to model the phonotactics property of the words in a language and data scrapping and cleaning to prepare the data sets. Stack: PyTorch (on Jupyter notebooks for the data preparation), for the app proper: Firebase (+ TS for the functions) and Vue3 + TS.
Research, experimentation and contribution to open source projects with minority languages to explore complex NLP challenges in low-resource contexts.

I work at the intersection of AI, linguistics, pedagogy, and programming. My work spans executive AI advisory, applied AI engineering, and the development of intelligent learning systems. I combine technical expertise with a deep understanding of cognitive processes and language structure.
For consulting, collaboration, or speaking engagements, reach out here.