Turning data into cosmetic innovation

"In the globalized beauty industry, L'Oréal must innovate ever faster to meet the desires and needs of customers looking for new products and services that respect their bodies and the environment. Our vision is to deliver services to our businesses that they have not yet considered," said Philippe Benivay, IS Experimental Data Intelligence at L'Oréal. "Data and artificial intelligence allow us to move faster to create cosmetic products that meet the infinite diversity of beauty needs and desires of consumers around the world. Talend, in our private cloud environment on Microsoft Azure, helps us transform R&I into a value-added service that can leverage all research data to deliver the best in cosmetic innovation in terms of quality, efficiency, and safety."

L'Oréal's R&I division deploys its teams around the world to imagine and create innovative products of impeccable quality. Every year, its development teams introduce several thousand formulas. These continuous advancements are the result of an ongoing dialogue between science and marketing. L'Oréal scientifically and rigorously demonstrates the safety and effectiveness of its innovations. To bring new products to market, R&I must be able to compile all data related to the characterization and physicochemical definition of formulas and raw materials, and the perceived performance of products by consumers in real conditions and in real-time. 

To have a more accurate view of formulas and product perception, L'Oréal wanted to capitalize and centralize all research data of various natures and formats. This need to integrate multiple data sources led to the selection of Talend and the implementation of a new data lake in a private IaaS on Microsoft Azure.

By deploying Talend Data Fabric, L'Oréal is able to connect all types of databases, structured laboratory data, and very heterogeneous and sometimes raw data sources, such as robotic measurement or images data. Talend's solution allows L'Oréal to incorporate intelligent algorithms directly into data integration flows in the form of APIs. The data lake enables the R&I to consolidate and prepare data to facilitate researchers' analyses and help them base their conclusions on reliable, high-quality data.

The data lake processes 50 million pieces of data per day. With Talend, data is available in real-time, and the data lake is refreshed several times a day, allowing more precise management of sensitive research projects. The first operational use case developed for the Finance Department was on the economic management of research. The new dashboards managed all KPIs for research-related activities and associated costs, such as tests for certification and qualification.

Since the commissioning of the data lake, use cases have multiplied at L'Oréal. To meet business needs, the company has deployed data marts on-demand to produce even more knowledge on many subjects, such as the microbiome (micro-organisms living inside and on the surface of the epidermis) or the exposome (the effect of pollution on the skin). Toxicologists have simplified, autonomous, and rapid access to quality data to support the development of new formulas. For consumer evaluations, all the videos that will be acquired during testing sessions to evaluate the effectiveness of the products are integrated into the data lake, allowing for almost instantaneous image processing and manipulation.

"Research teams must base their analyses and conclusions on trusted data that is accessible in real-time," said Ciaran Dynes, SVP of Products, Talend. "L'Oréal's R&I team uses data and artificial intelligence not only to help them develop new and increasingly innovative products, but also to inform other departments throughout the company. At Talend, we are proud to be able to participate in the innovation of the world leader in beauty."