Beiersdorf’s team of over 900 scientists has been pioneering innovative skincare solutions for decades. By centralizing their vast data library and utilizing AI-powered knowledge mining through Microsoft Azure Cognitive Search, researchers can swiftly access the information they need. Features like document summarization, semantic search, document tagging, and optical character recognition enable quick extraction of key information, allowing researchers to focus more on leading product development rather than time-consuming searches.
Our in-house Azure-based search engine is equipped with AI and semantic capabilities to index terabytes of data across various heterogeneous data sources—thereby democratizing Beiersdorf’s R&D knowledge among our research scientists.
Stephan Abend: Head of the Data Science Hub
Beiersdorf
Decades of research and skin care innovationÂ
Beiersdorf has been producing superior skin care products since 1882, including leading brands such as NIVEA, Eucerin, La Prairie, Labello, and Hansaplast. For decades, the company has been rigorously testing, analyzing, and evaluating its products to maintain their excellence, resulting in a comprehensive library of institutional knowledge. Beiersdorf’s team of 900 scientists relies on this library to innovate and expedite their work. However, finding the necessary information has been challenging due to the spread of data across more than 800 SharePoint sites and Beiersdorf’s Data Lake. To address this, Beiersdorf enhanced its in-house R&D search engine with Microsoft Azure Cognitive Search and the document summarization feature in Azure Cognitive Services for Language.
Building a centralized, searchable digital libraryÂ
To ensure its digital transformation efforts meet the centralization, organization, and accessibility needs of its researchers, Beiersdorf assembled a cross-functional team comprising data engineers, data scientists, machine learning engineers, and full-stack developers. This team concentrated on unifying Beiersdorf's resources and developing tools to empower R&D teams to continue delivering cutting-edge research and skincare innovations.
Stephan Abend, head of the Data Science Hub at Beiersdorf, is a mathematician with experience in creating data-driven solutions across various industries, including banking, media, and FMCG. He initiated the data transformation efforts at Beiersdorf. "Tech is the infrastructure, the vehicle, that allows us to bring real-world products into a digital realm where they can be processed accurately and algorithmically," says Abend. "To truly make data-driven decisions, you need digital solutions."
To create a searchable, centralized database, Beiersdorf implemented Azure Cognitive Search, enabling researchers to obtain precise, direct answers from multiple data sources. Leveraging semantic search and document summarization, researchers receive AI-powered results that contextualize, summarize, and highlight the most relevant information. "Our in-house Azure-based search engine, equipped with AI and semantic capabilities, indexes terabytes of data from various heterogeneous sources, thereby democratizing Beiersdorf’s R&D knowledge among our research scientists," explains Abend.
Streamlined access to institutional knowledge
Beiersdorf has developed an advanced AI-powered platform that leverages the company's resources through seamless, user-friendly features and custom-built solutions. The platform efficiently extracts relevant information from lengthy documents utilizing document summarization and keyphrase extraction from Azure Cognitive Service for Language, providing context and condensing essential knowledge.
Lengthy paragraphs are distilled into sentences with document summarization, and pages of irrelevant information are transformed into concise outlines. This significantly reduces the time researchers spend scanning texts, helping them navigate vast amounts of data. Additionally, optical character recognition (OCR) enables quick access to scanned documents and hard-to-read PDFs by converting printed or handwritten words into text.
Looking forward, Beiersdorf plans to extend these search and data capabilities to other departments across the organization. By applying AI data mining technology, employees will gain deeper customer insights, analyze feedback, and better understand consumer perception.
Tailored solutions for industry-leading innovation
Several customized features adapt Beiersdorf’s solutions to the skincare industry. For instance, a custom-built document tagging feature enables researchers to categorize research by topic or product area. Leveraging the semantic search feature of Cognitive Search, Beiersdorf employs machine learning to group words and phrases specific to the skincare industry. For example, a search for "ascorbic acid" will also retrieve studies on vitamin C (another name for the component). This ensures researchers find all relevant information without wasting time searching for different phrases or variations of their query.
We’re able to free our researchers from interacting with dozens of different interfaces, which means more quality time spent on research and innovation.
Stephan Abend: Head of the Data Science Hub
Beiersdorf
Less searching, more scienceÂ
For over 140 years, Beiersdorf's research has enabled the creation of products that support people's daily self-care—a legacy that deserves to be fully utilized. Azure AI empowers R&D teams to focus on their core strength: research.
"Beiersdorf aims to hire the smartest brains in biochemistry," says Abend. "By freeing our researchers from interacting with numerous interfaces, we ensure more quality time is spent on research and innovation."
The technology infrastructure ensures that new research will be dynamically archived, tagged, and summarized, guaranteeing access for future teams. With these capabilities, Beiersdorf can leverage its research strategically and effectively, maintaining its leadership in superior skincare solutions.