I'm a Software Engineer and recent graduate with a dual Master's degree in Computer Science from Polytech Lille and Université Laval. I enjoy designing robust software architectures and integrating AI into practical applications.

My experience spans R&D, AI model evaluation, and data optimization:

• Evident Scientific: developed an AI model evaluation module from end to end for segmentation and anomaly detection, while substantially reducing RAM and VRAM usage.
• Acteam-IT, working with Mondial Relay: designed cost optimization solutions and AI data generation pipelines deployed internationally.

I also build complex open-source projects, including Meridian, a graph-based AI workflow platform built with Python and Nuxt, and Go tools such as Spritely and gouv.viz.

A graph-based AI workspace for building visual workflows and branching conversations. Connect models, tools, and data sources, with parallel model queries and a chat interface. Built with Nuxt and Python.

Sync Letterboxd watchlists with Radarr and Jellyfin collections, with optional Sonarr support for series. Automate movie discovery and remove watched films from your Jellyfin watchlist collection. Built with Python and deployable with Docker.

A French cinema planner with visual showtime timelines, search by available time window, and direct booking links. Compare screenings across cinemas and find a film that fits your schedule. Built with Nuxt, Go, and PostgreSQL.

An open-source Android manga reader connecting Komikku's reading experience to a Suwayomi server. Browse your server library, manage sources, and download chapters to your phone for offline reading, with progress synced when you reconnect. An independent fork of Komikku.