While everyone is trying to explore how to apply artificial intelligence to various industries, French startup OsiumAI has discovered an interesting application case for artificial intelligence - the research and development of materials science.
Founded by Sarah Najmark and Luisa Bouneder, the startup raised $2.6 million in seed funding from Y Combinator, Singular, Kima Ventures, Collaborative Fund, Raise Phiture and several angel investors (Julien Chaumond, Thomas Clozel, Isaac Oates, Liz Wessel, Ebert Hera Group, Patrick Joubert, Sequoia Scout and Atomico Angel).
During my undergraduate studies, I worked on materials, particularly in the field of cosmetics. Sarah Najmark told me: "I see that materials development methods are still very manual, involving a lot of trial and error, and many methods rely mainly on intuition."
After graduation, she joined GoogleX, the giant tech company's moonshot arm, and spent three years studying robotics and deep technology. She has co-authored several patents.
"As a technical lead I really own an end-to-end AI pipeline for robotics and systems engineering topics," she said. "In discussions with many industrial companies, we are also aware of new challenges related to sustainability, namely the development of new materials: lighter materials (such as aerospace materials), but also more durable and environmentally friendly materials, as well as optimized and greener manufacturing processes." "
Co-founder Luisa Bouneder has spent three years developing data products for industrial companies, particularly in the materials field. She also noticed that a lot of trial and error slowed down the development process. "This is a topic that really affects a variety of industries, including construction, packaging, aerospace, textiles and smartphones," she commented.
So, how exactly does OsiumAI work? This is the feedback loop between optimizing material formulation and testing using a data-driven approach. Using the startup's proprietary technology, industrial companies can predict the physical properties of new materials based on a range of criteria. OsiumAI can then help refine and optimize these new materials while avoiding common mistakes in the trial and error process.
Several industrial companies are already trialling OsiumAI’s solutions and see the potential. "Our users believe that our solution allows them to develop and analyze materials 10 times faster. So from the very beginning of testing, we saw the value we bring," said Najmark.
In many ways, OsiumAI is just getting started. The company currently only has two employees (two co-founders), so the startup will soon expand the team and start converting these industrial tests into formal contracts.