Dr. Mohamed Aouad

Why independent validation

In every technical project I have been close to, the expensive mistakes were the ones nobody outside the team had checked: a model that assumed a parameter the datasheet contradicted, a backtest that flattered a strategy, a design that could not reach its target even in the best case. The fix is cheap and early, and it is almost never bought, because the people who could do the check are either inside the project or trying to sell the next phase of it.

ALPHA-AQAL exists to be the outside check. I have no prototype to build for you afterwards and no stake in the round. I read the physics, run the code, and write down what holds.

The career

I trained as a materials engineer at École Polytechnique d'Alger and as a physicist at Université de Lorraine, with research in plasmonics at Institut Jean Lamour in Nancy. My doctorate at CEA-Leti and Université Grenoble Alpes was on the physics and compact modelling of FDSOI transistors down to cryogenic temperatures, in the team working on the quantum-computer project. That work led to an invited presentation at the MOS-AK workshop in Grenoble and to reviewing for Solid-State Electronics.

After the PhD I joined Erdyn, a French innovation consultancy, where I ran technology and market studies for research institutions, technology-transfer offices and industry, and led an internal programme on using language models in consulting work. Since 2025 I have worked as an independent scientific consultant: thirteen completed client missions, from quantum devices and magnet design to clinical study feasibility, trading systems and medical AI. I also trained as a full-stack developer, which is why the analysis pipelines behind this practice are built and maintained in-house.

How I work

Remotely, in writing, and with a gate on every step. Specialised AI agents do the legwork of literature, calculation and simulation; I verify each step before it goes further and send back what does not hold. Every report ships with the scripts and inputs that reproduce it, because a verdict you cannot check is just an opinion.

What I believe

• Evidence over opinion. A number with its source beats an adjective.
• A no-go delivered early is a result, and it is worth exactly as much as a go.
• If a claim can only be settled on a bench, say so and name the test, rather than guess.
• Show the working. Anyone the client chooses should be able to reproduce the report.