From Design to AI-Powered
Software Production

Before any production, I implement a phase of reflection and analysis to understand the need and build a response aligned with the specifications. This stage also includes analysis of the existing landscape and literature for topics with an R&D dimension. It is in this phase that accumulated experience takes on its full meaning. It guides technical and methodological choices, allows anticipation of constraints, and secures the chosen directions. These foundations then ensure efficient, coherent, and sustainable code and documentation production.

For production, I use a set of tools and technologies adapted to the various needs encountered during a software project. Some are part of my daily work environment, while others are mobilized in a more targeted manner depending on the technical context, performance constraints, or application domain. All technologies presented below have been implemented on concrete projects, in real-world situations, whether for application development, 3D visualization, data processing, automation, or deployment. This technical foundation allows me to tackle various challenges while ensuring coherence, reliability, and maintainability.

I now integrate Claude Code at the heart of my workflow as a true productivity lever. Having held the role of project manager in the past, I find in this new paradigm a familiar logic: steering, framing, arbitrating, and validating — except that I now address coding agents. When data confidentiality is a concern, I switch to a model operating locally, without any transmission to external services. This synergy between human expertise and AI assistance gives me a production capacity equivalent to that of a small team, with the coherence and accountability of a single point of contact.