Entalpic

Entalpic is a French deeptech company that applies artificial intelligence and atomistic simulation to the discovery and engineering of molecules and materials for surface-driven industrial processes. Its platform, described as a high-throughput or atomistic discovery engine, screens large chemical spaces and ranks candidate materials under industrial constraints, combining generative models with predictive workflows built on physics-based simulations and machine learning. The company also models reactors and experimental systems using machine learning combined with computational fluid dynamics (described as digital twins/process modeling), so that candidate materials can be assessed under manufacturing conditions.

The underlying models draw on multimodal datasets: quantum/atomistic simulation data from open repositories and proprietary high-throughput pipelines, structured analysis of patents and academic literature, and experimental data produced with partner laboratories and, prospectively, an in-house lab. Reported technical building blocks include graph neural networks, GFlowNets, active learning, and large language models used for mining scientific literature. Application areas cited include catalysis (notably ammonia synthesis catalysts), battery materials and cathode/interfacial chemistries, semiconductor thin films and coatings, advanced materials and precursors for deposition processes, carbon capture, and green hydrogen.

Entalpic operates from Station F in Paris with a presence at Mila in Montréal. Its team is described as machine learning, chemistry and materials specialists, more than half holding PhDs, with alumni of Intel, Air Liquide, Applied Materials, Meta, Google and Amazon.

Founding story

Entalpic was founded in 2024 by two Mila researchers, Alexandre Duval and Victor Schmidt, together with Mathieu Galtier, a former executive at Owkin. Duval and Schmidt's work originated in Yoshua Bengio's and David Rolnick's labs at Mila, where they published on predictive and generative models for materials and catalysis modeling. Duval completed a PhD from CentraleSupélec and Inria Saclay in December 2023 while attached to Mila's 'AI for Climate Change' group, and previously worked at Amazon; Galtier served as Chief Data & Platform Officer at Owkin, where he built industrial data partnerships. Galtier is CEO, Schmidt CTO and Duval CSO [2][6][7].

Business model

Entalpic combines open and proprietary research, positioning its data assets — quantum simulations, lab experiments, literature and patents — as a competitive edge. Reported commercial approach centers on collaborative and co-development projects with industrial partners and academic labs, with intellectual property generated through co-patented discoveries [2][6].

Sources describe an IP- and partnership-based model (co-development projects and co-patented discoveries with industrial clients) rather than a disclosed pricing or subscription structure [2][6].

Traction

Publicly reported traction is early-stage: a €8.5M seed round in 2024, a stated team of 25 in 2025 with an end-to-end technical proof-point, collaborations with industry partners and academic labs, and advisory support reported from Yoshua Bengio and Hugging Face co-founder Thomas Wolf [0][4][5][6]. As of the Mila interview the company was focused on industrializing its platform and had not yet reported validated commercial deployments [2].

Latest developments

Company communications dated 2025 describe applications extended beyond catalysis to semiconductors, batteries and advanced materials, an emphasis on surface chemistry and solid-fluid interfaces, and planned milestones of commercial expansion in early 2026 and a Grenoble experimental lab in late 2026 [0].