Curiosity
Questions come first. We follow them across disciplines, from how models behave to how systems are put together.
AI & software engineering
ZEXSSO — Aix-en-Provence, France
An engineering company building agentic software, intelligent systems, and the technologies shaping what comes next.
01About
ZEXSSO is an engineering company based in Aix-en-Provence. We build agentic software and intelligent systems — patiently, precisely, and with a clear view of what they should become.
Questions come first. We follow them across disciplines, from how models behave to how systems are put together.
Ideas are cheap until they are tested. We prototype early, measure honestly, and keep what holds up.
Details compound. Careful engineering is what lets an ambitious idea survive contact with reality.
02Focus
Four disciplines we keep returning to. These are fields of interest and inquiry — the questions that shape our work.
Exploring language models, intelligent systems, and new approaches to human-computer interaction.
How models reason, where they fail, and what interfaces make them genuinely useful.
Designing robust, efficient, and maintainable software architectures.
Clear boundaries, explicit data flow, and systems that stay legible as they grow.
Investigating adaptive workflows, agentic systems, and programmable intelligence.
Agents that act within well-defined limits, and the tooling needed to trust them.
Experimenting with emerging technologies and turning technical ideas into practical systems.
Small prototypes, honest measurements, and a short path from idea to evidence.
03Approach
A small set of habits, applied consistently. They are less a doctrine than a way of keeping our own work honest.
Start from the problem, not from the trend. Form a view, then test it against reality.
Most constraints are inherited. Find out which ones are real before building around them.
Every dependency, abstraction, and line of code should earn its place.
Simple systems are easier to reason about, to change, and to rely on. Complexity is a cost, not a credential.
Keep room for experiments that may not work. That is where the interesting answers tend to be.
04Location
Based in the south of France, between the Mediterranean and the limestone ridge of Sainte-Victoire. A good place to think carefully.
Aix
05Contact
For research conversations, collaboration, or simply to compare notes on a technical idea — write to us.