3 questions to Arnault IOUALALEN, CEO to Numalis
Le 06/09/2021
Arnault IOUALALEN – CEO – Numalis
Present your company
Numalis is a software editing company that provides tools and services to help with the design and validation of AI systems. Numalis accelerates the conception, training and validation phases of AI. To do this, Numalis uses its expertise in AI analysis considering of formal methods for the measure of their robustness and explainability. As such, Numalis is involved in ISO/IEC standardization committees, AFNOR and French High Health Autority aiming to make AI more reliable and more explicable. Mr. IOUALALEN, CEO of Numalis, is editor of the ISO 24029 series of standards on the topics of AI robustness.
What are you going to present at the first edition of SIDO Paris?
Saimple is a software helping data scientists and AI engineers to build more reliable and efficient AI models faster. Machine Learning and Deep Learning design, training and validation phases are accelerated with our solution, and the reliability of the final AI model is increased. Saimple uses formal methods to validate the robustness and provide explainability of neural networks. Our solution is available as a SaaS or on premise. For more information, visit our website : https://numalis.com/
How does your product / solution contribute to your industry?
Based on its unique know-how in the validation of critical numerical embedded systems, Numalis industrializes more than 25 years of R&D that allow it to develop very first SaaS tools to validate AI systems with its product: Saimple. The software relies on formal methods to assess the robustness and explainability of neural networks, helping to build more reliable and efficient. AI models faster. Saimple brings right from the start of a project a unique value to both AI engineers and stakeholders for the quality assessment of the AI product. Mitigating AI risks is essential to the success of any AI product. Incorrect mitigation can have dramatic cost depending on when the risks manifest themselves. Estimated upstream risks (R&D, prototype) of AI within $100K and $1M ; downstream risks (integration, deployment) of AI within $1M and $10M ; in-operation risks of AI within $10M and $1B.
Numalis will be on SIDO Paris on the BOOTH J23 |
Learn more about Numalis :
Website : https://numalis.com/