About us

Our goal is not to mimic human intelligence, but rather to augment it through the use of long term prediction tools
Classical AI
Qualitative answers (mimic human intelligence):
classification and pattern recognition
(the glass is half full)
natural language processing
Need to tune many parameters to get satisfactory results
Big data oriented
Our approach
Deep Quantitative Learning (augmented human intelligence):
description of the system with high precision
(the glass is 52.3578 % full)
long term prediction of continuous phenomena
Fully autonomous learning process with no risk of overfitting
Efficient even when the amount of data seems limited
Create Digital twin of your system

Digital twin is a dynamic digital representation of a physical object or process.
Using data from multiple sources, the digital twin continuously learns and updates to reflect any change to the physical counterpart.

You can test your ideas on the digital copy before applying only the best one to the real system.

  • Learning real-world phenomena from data
  • Long-term prediction for dynamical systems
  • Highly nonlinear compression
  • Accurate description of the system, instead of just yes/no answers regarding its state

In these domains, classical deep learning methods are clearly failing. We had to create groundbreaking learning methods to achieve these objectives.

The parsimonious and topologically optimized architectures of our solutions guarantee satisfactory results even when the dataset is small compared to the complexity of the sytem (which is often the case) and get rid of overfitting risks.

Example of network architecture

Deep Quantitative Learner Framework


deepROM® is our neural network factory. It excels at figuring out automatically the optimal network topology. The whole process is then standalone and avoids overlearning. Hence the user only has to provide the learning data and shortly after gets a reliable model. See examples here.


xROM™ is our implementation of recurrent reduced order models. It builds models for nonlinear dynamic processes and get rid of their inherent instabilities. See examples here.


Our convolutional neural network solution is adapted to higher dimensional problems. It gets rid of the so-called curse of dimensionality. It controls adaptively the data generation process. See examples here.

Our main applications

Our partners

What our clients say

En rupture totale par rapport à l’état de l’art, la société ADAGOS arrive à créer des modèles réduits ayant un grand nombre de paramètres (plusieurs centaines dans les cas considérés). Cela veut dire que ces outils ne semblent pas être affectés par ce que l’on appelle la malédiction de la dimension (curse of dimensionality). Cela parait d’autant plus étonnant que nos paramètres varient dans des intervalles très larges et ont un caractère discret aussi bien qualitatif que quantitatif.

Jean-Marie Hamy Technical Project Manager and Design Authority @ AREVA NP

Les enjeux sont considérables pour les systèmes embarqués et les applications médicales. De ce fait, nous maintenons une veille technologique et nous sommes en contact avec les universités les plus prestigieuses pour scruter la moindre évolution dans le domaine. Mais, il n’était pas nécessaire d’aller chercher bien loin, l’innovation de rupture peut venir d’une petite startup toulousaine appelée ADAGOS.

Michel Rochette Director of research @ Ansys France

Our team

9 avenue de l'Europe - 31520 Ramonville-Saint-Agne