Skip to content

Ahlame Douzal’s Interview in L'Usine Nouvelle

xpdeep in l'usine nouvelle : oui au deep learning, non à la boîte noire


Yes to deep learning, no to the black box: how the deeptech Xpdeep produces self-explanaible neural networks.

This is the title of the article in L'Usine Nouvelle that has captured the attention of French industrials: an in-depth discussion with Ahlame Douzal, our Chief Scientific Officer, about our pioneering self-explanaible deep learning framework. The interview sheds light on how this innovation propels AI into a new realm of transparency and accessibility, marking a thrilling advance in the field.

In an industry where AI has often been likened to a 'black box', our company, Xpdeep, stands out by producing neural networks that articulate their internal decision-making processes. This not only demystifies AI but also propels its adoption across industries wary of the opaque nature of conventional AI technologies.


Douzal's insights in L'Usine Nouvelle – a publication esteemed for its industrial acumen – underscore the potential of our technology to be a game-changer. Scheduled for commercial release in January 2024, our platform promises to illuminate the 'black box' of AI, delivering comprehensible and actionable insights.

Backed by scientific research and a collaboration born out of the Grenoble Informatics Laboratory (LIG), Xpdeep's approach goes beyond mere performance. As Douzal eloquently puts it, "Performance alone is not enough when it touches the health of people." This philosophy is reflected in our software's capability to decipher the intricate web of decisions made by AI, providing a detailed narrative of its reasoning.


"Performance alone is not enough when it touches the health of people."

photo-ahlame carre-xpdeep
Ahlame Douzal, Chief Scientific Officer and Co-Founder


Real-World Applications

The implications of our technology are vast, with a demonstrated success rate of 93% in identifying human activities through multivariate time series. But what sets Xpdeep apart is the resulting decision graph, which intricately maps out the AI's learned decisions, offering clarity on discriminants in data segmentation – a feature that's already piqued the interest of the automotive industry.

As we gear up for the launch of our self-explanatory deep learning framework, the industry watches with keen interest, anticipating a new era of transparent AI.

To learn more about our transformative AI solutions and stay updated on our upcoming releases, subscribe to our newsletter !