Behind the Term Sheet: Entalpic’s $8.5M Seed
September 11, 2024
Europe
The Paris-based startup decarbonizing industrial chemistry with AI
AI is transforming industries at an unprecedented pace — AI-driven drug discovery, fraud detection, quality control, personalized commerce, autonomous vehicles, smart grids, the list goes on.
But in the last few years, we’ve seen the emergence of AI being applied to a new field with great promise: new materials discovery. It’s attracted some of the world’s top researchers with new projects and publications such as Google DeepMind predicting 2.2M new materials.
Major corporations across industries are also taking note. Similar to the success of AI in drug design, accelerating the discovery of novel materials with tailored properties (e.g., enhanced strength, conductivity, thermal resistance, or biocompatibility) opens the door to a world of breakthroughs and possibilities — in energy or carbon capture to electronics and the chemicals industry.
This is what led us to our latest investment in Entalpic — co-leading its €8.5M seed round alongside Breega and Felicis. This startup is at the forefront of generative AI technology for the chemical industry with the mission to decarbonize the sector. It was founded in 2024 by researchers out of Mila, the premier machine learning lab for climate action, and an early employee at Owkin — a leader in AI for drug discovery and a Cathay Innovation portfolio company we’ve supported since 2018.
Blending open and proprietary research, the company is building an advanced AI platform that designs catalysts (substances that increase the rate of a chemical reaction) to optimize chemical processes in areas like energy storage, fertilizer production and pollution control.
Here’s why we believe Entalpic is set to transform the chemical industry.
FIRST, THE MARKET — CRYSTALLOGRAPHY MEETS AI
Crystallography is the cornerstone of material design, drug development and cutting-edge technology. By delving into the atomic arrangements of materials, scientists have been able to unlock critical insights into properties like strength, electrical conductivity and chemical reactivity.
However, the traditional path to discovering novel crystal structures is anything but straightforward. It relies on a time-consuming, trial-and-error approach — tweaking known crystals or experimenting with new elemental combinations — often requiring months of painstaking effort to yield limited results. The field has leaned heavily on physics-based methods like Density Functional Theory (DFT) for predicting material properties. While DFT is undoubtedly effective, it’s also resource-intensive, demanding significant computational power and time.
But with the advent of AI, particularly machine learning, the crystallography market is undergoing a paradigm shift. In the last year, we’ve seen a surge of research publications from the likes of Google, Meta and Microsoft that highlight the use of AI for breakthroughs in crystallography.
A standout development is the application of Graph Neural Networks (GNNs) combined with quantum theory for the generation of new crystal structures. The traditional approach typically started with known structures and worked forward. But in this new AI-driven paradigm, the process flips on its head: it begins with the desired material properties and works backward to identify potential candidates. These candidates are then put through rigorous validation via DFT before moving on to synthesis and testing in the lab.
Moreover, a new frontier is opening up in the application of large language models (LLMs) to the synthesis process, helping researchers discover novel recipes for material synthesis. By integrating AI into every stage of crystallography — from structure generation to final synthesis — this approach promises to unlock unprecedented innovation, making material discovery faster, more efficient and more tailored to specific needs.
Overall, the transition to AI-driven crystallography is not only accelerating the pace of discovery but also expanding the possibilities for innovation in material science — unlocking new potential in material design, drug development and beyond.
ENTER ENTALPIC — BRINGING GENERATIVE AI TO INDUSTRIAL CHEMISTRY
Entalpic founders Victor Schmidt, Mathieu Galtier, Alexandre Duval
Entalpic is building the fundamental generative AI model specifically designed for materials discovery in crystallography and the large, carbon-intensive chemicals industry.
Leveraging the latest in AI technology, including LLMs, Active Learning and Graph Neural Networks — its platform generates and evaluates new materials and molecules to replace outdated industrial chemical processes. It formulates and tests chemical hypotheses through automated experiments, enriching the knowledge base with actionable data — all at the speed of AI.
The Highlights:
The Large Market Opportunity — focusing first on industrial chemistry. By starting with catalysts for ammonia production (critical for fertilizers), Entalpic is tapping a $100B high-impact market, operating on a single, century-old process that accounts for more than 1% of global CO2 emissions. Further, all carbon intensive industries rely on catalysts with the potential to improve efficiency, reduce costs and enable novel processes to address global environmental challenges.
Examples: Chemicals (polymers, fertilizers, specialty chemicals), pharma (drug discovery and development); electronics (semiconductors), automotive (reduce emissions of harmful pollutants from vehicles), energy (cleantech, hydrogen, carbon capture etc.)
The Business Model — open & proprietary research enhances data edge. This includes catalyst datasets from diverse sources from numerical quantum simulations and physical lab experiments to academic literature and patents. These datasets not only feed its generative model, but the approach fosters greater collaboration in the industry on initiatives such as green hydrogen production while advancing in-house developments with industrial clients.
The Founders — the Owkin playbook combined with deep AI expertise in materials science. Mathieu Galtier (CEO), was an early employee at Owkin, the billion dollar startup applying AI to drug discovery. As Chief Data & Platform officer, he developed the company’s data partnership network with industrial leaders — minting strategic partnerships that granted access to proprietary datasets and opportunities that led to the co-development of new AI therapeutics applications in various disease areas. Victor Schmidt (CTO) and Alexandre Duval (CSO) hail from the world renowned AI research lab Mila (founded by Yoshua Bengio) where they authored a series of papers on predictive and generative models for materials and propose several model-agnostic architectural improvements for catalysis modelling.
Combining these two areas of expertise enables Entalpic to become a leader in GenAI for materials discovery and a trusted partner to leading academics and corporations in the co-development of new materials that address critical pain points — accelerating adoption in industry practices and the decarbonization of processes.
PARTING THOUGHTS — FRENCHTECH AI FOR A SUSTAINABLE FUTURE
The Entalpic story is one that closely aligns to us at Cathay Innovation: We are a French Venture Capital firm with a global platform spanning Europe, the US and Asia. We strongly support the French AI revolution and have dedicated resources to back the startups and entrepreneurs pushing it forward (more on our Frenchtech AI initiative here). We’re big believers in industry collaboration — so much so, that we bake it into our model with our corporate ecosystem that are not only investors, but potential partners to our portfolio companies. And as early investors in Owkin, we’ve done this before.
The application of AI to new materials discovery has immense potential — expanding the possibilities for innovation in material science and the decarbonization of industries — with Entalpic well positioned to become a leader. We’re thrilled to join the Entalpic journey in its early innings as it sets its sights on industry partnerships and expanding operations (currently in France, Germany and Canada) to the US and Asia. With its game-changing technology, business model and highly skilled team, we’re confident that this startup will redefine material and molecule discovery for a more advanced and sustainable future.
Written by Paul Pouyanne & Jacky Abitbol; Edited by Valentine Dufeigneux & Jaclyn Hartnett