Atomistic AI for Materials Engineering

Alexandre Duval, Co-Founder and CSO, Entalpic

© Entalpic

Materials underpin every modern technology, yet R&D is still constrained by human bandwidth: too many possible formulations, too much scattered information, and too many objectives to optimize at once, with limited access to rapid experimentation. Entalpic built a materials discovery platform that integrates AI models, automated quantum simulation workflows, and experimental validation. CHEManager spoke with Alexandre Duval, CSO and co-founder, about turning AI insights into deployable materials.

CHEManager: What is Entalpic’s core idea in one sentence?

Alexandre Duval: We replace trial-and-error materials R&D with an AI closed-loop discovery engine that turns millions of candidates into a few lab-validated materials, ready for industrial deployment.

What makes R&D for materials discovery so slow in the chemical industry?

A. Duval: Materials R&D is slow because chemists rely on human intuition to find a new material within an infinite search space, using slow, expensive, and often hard to reproduce lab experiments to validate their hypotheses. On top of that, teams need to optimize for multiple desirable properties under many constraints, accounting for years of prior research, which makes progress hard to scale without better tools.

Many companies claim “AI for materials.” What is different about your approach, and what chemistries do you focus on?