AtelierMoléculaire

AI-driven molecular innovation for beauty brands.

01What

AI is making exponential progress in technical fields including industrial engineering, materials discovery, and biopharma. We bring these capabilities to beauty brands to radically alter the technical innovation cycle: new and better molecules and formulations, engineered faster and with greater precision.

Early-stage drug discovery is being transformed by rapid progress in AI. This is particularly true of molecule design, where models are now powerful enough to reason about molecular structures. The bottleneck is moving to experimentation: in response, lab-in-the-loop, where AI design drives in vitro experiments, is emerging as a potent format to rapidly engineer better pharmaceutical molecules.

We adapt the cutting edge of molecule discovery to the consumer space. Beauty brands are at the forefront of the mounting consumer demand for newer, more personalized molecules that better connect wellness to health. We are building the platform to meet this demand.

02How

We partner with beauty brands to build new molecules and formulations that meet their objectives, with increasingly higher performance.

We build a lab-in-the-loop for beauty, so that we can best discover new peptides and molecules with optimal lab data, both in terms of safety and efficacy. By combining AI design with in-house experiments, we explore vastly more design space than possible otherwise, guided by concrete lab data.

Example We might engineer peptides or small molecules with high performance on skin clarification. We would rapidly iterate over hundreds of designs, feeding lab data back to our design engine in iterative fashion, resulting in molecules that meet the objectives of our partner.
Lab-in-the-loop, schematically: each round, new peptide designs are tested on cells, and the readout guides the next round.

03Who

Founding team with deep expertise in AI, engineering, molecular design, consumer goods, and beauty. Serial entrepreneurs with several exits and long scientific, operational, and business track records.

04When

Now. We expect to work with a small number of design partners as we build our platform.