I am curious about how AI will change the way we work, and how more people can benefit from it. I began my PhD in late 2022, shortly before ChatGPT was released. Since then, I have explored applied AI through research, internal tools, products, and a mobile app.
What interests me most is the point where an idea becomes a system someone can use. At DxO, my team and I build and evaluate AI features and rethink the workflows around them. I am particularly interested in AI as a new interface for software: a person expresses an intention, while an agent works through explicit tools and leaves the result editable. Silver AI is where our team is exploring that idea for photography.
Previously: mathematics and machine learning at ENS Paris-Saclay, followed by remote-sensing work at Kayrros, CNRS, and Inria. I recently wrote up part of that research as a France-wide comparison of AlphaEarth embeddings and hand-built satellite features.
Nik Silver Efex is a reference for professional black-and-white photo editing. Silver AI turns a photographer’s intent into visible, editable actions inside the software. I lead the DxO team preparing the research preview.
Drawing on public accounts from teams at OpenAI and Anthropic, I set up a Linear-based workflow in my team. It connects human decisions, bounded agent work, pr-preflight, CI, and review. I now advise other R&D teams as they adapt it.
Outside DxO, I co-founded DailySurf and serve as CTO. I lead product engineering and have built most of the app, which helps surfers record sessions, remember conditions, and follow their progress.
A conversation with I-Academy about introducing AI into day-to-day work across a company. In French.
My speech at the 2023 ENS Paris-Saclay graduation ceremony begins at 1:44:00. In French.
AI made the app faster to build. It did not make finding the right product easier.
Nik Silver Efex is a reference for professional black-and-white editing. Silver AI turns a photographer’s intent into visible, editable actions inside the software.
How I use Linear, bounded agent work, pr-preflight, and human review in one engineering workflow.
Smartphones, AI, and the fact that products evolve faster than people do.
A personal account of leaving academia and choosing to work on products.
A controlled France-wide comparison of AlphaEarth embeddings and hand-built Sentinel-2 phenology features.
An unpublished framework for reconciling incomplete event records across space, time, and source reliability.
A compact 1.6-million-parameter U-Net trained to remain useful across dates, seasons, and biomes.