About the Webinar

What if we could capture the true sensory experience of a skincare product—without relying on written questionnaires?

In this webinar, we will explore how AI-powered analysis of body language, non-verbal cues, and vocal responses extracted from recorded videos of volunteers assessing different products can reshape sensory evaluation in cosmetics. By analyzing unconscious physiological and expressive signals during real product application, this approach moves beyond traditional written assessments to reveal more authentic consumer perceptions.
​​​​​​​
We will explain how generative AI models were tailored through prompt engineering to assess verbal and non-verbal responses. We will compare these results to volunteer-reported ratings, an expert-annotated ground truth, and physico-mechanical characterization of the products. We will also demonstrate how this analysis enables the identification of outlier responses, uncovering individual sensory perceptions that are often masked in self-reported data.
The result of this proof of concept sets the path for a smarter, more objective way to decode sensory experience—opening new possibilities for designing skincare products that people instinctively connect with.

About the Speakers

Laura Molina holds a degree in Chemical Engineering and a master’s degree in Biological and Environmental Engineering.  With more than 10 years’ experience in material science and biotechnology, she has built her career as researcher and process developer for the cosmetic industry.  She has also participated in several funded European projects and helped small skin care focused finished good companies thrive. Since 2021, Laura works as Skin Care Applications Scientist at Lubrizol Advanced materials in Gavà (Spain). She is dedicated to introducing and delivering new concepts and ingredients to the skin care, sun care and color cosmetics market.

​​​​​​​Quinn Frank is a Data Scientist for the Lubrizol Corporation, based remotely in Raleigh, NC. He supports statistical modeling and analytics tools for specialty chemical R&D across multiple sectors, from engine oil additives to skin care products. Much of his work involves machine learning applications, especially in the computer vision space, as well as the software/data engineering necessary to facilitate this. He has a BS degree in Mathematics and Computer Science from Georgetown and completed a Master's in Statistical Science at Duke in 2021.