Developed in collaboration with Microsoft, this AI solution revolutionizes access to decades of scientific knowledge, boosting innovation and efficiency in dermatological treatments.
Almirall, a Spanish pharmaceutical company, transformed its research and development (R&D) process using Microsoft artificial intelligence. Using Azure Open AI, the company drastically reduced the time spent searching for information in 400,000 documents accumulated over more than 50 years of research. The solution enabled scientists to access historical data in seconds, eliminating processes that previously took hours or days, and increasing query efficiency by up to 80%.
Specializing in dermatology for over 80 years and with over 50 years of data, Almirall had a strong foundation to accelerate drug development, reduce attrition, and build on past research. But accessing this knowledge was slow and inefficient. Researchers spent hours, even days, searching through scattered files and outdated systems. Valuable insights were often hidden, lost to retired colleagues, or unknowingly duplicated, wasting time and resources.
To address this issue, Almirall needed a faster, more accurate way for its R&D teams to access data. Microsoft offered a comprehensive set of services that easily integrated with Almirall's systems. The Microsoft Industry Solutions Delivery (ISD) team worked closely with Almirall's scientists to co-design and deliver a customized, results-driven solution aligned with their needs to enable innovation at scale.
"Our professionals are scientists. They have a specific language and knowledge. The Microsoft Industry Solutions Delivery team had the knowledge and experience in pharmaceuticals to speak the same language. Together, we were able to quickly develop a solution that provides immediate value to our R&D team," says Francesc Catalan, CIO and Director of Corporate Information Technology at Almirall.
Almirall built a custom assistant using Microsoft Azure OpenAI in Foundry Models, Azure AI Search, and Azure Databricks to help researchers quickly find answers among 400,000 documents in English, Spanish, and Catalan. R&D team members refined the assistant by testing commands and validating results. Now, scientists can locate past experiments and scientific papers in seconds, freeing up time for more innovative and in-depth work.
Almirall has since implemented the assistant across R&D teams, aiming to cost-effectively expand project architecture and internal knowledge to other departments.
Faster responses and treatments
Users are already noticing productivity gains, reporting finding accurate answers to their questions about 80% of the time. “When we had an issue with one compound, I asked the [AI] assistant if we'd seen a similar issue with another compound. It found examples in minutes,” says Montse Erra, Director of Discovery and Early Development for NCE at Almirall. “Now I can spend my time on what really matters, not half the day researching.”
In addition to the custom-built assistant, Almirall is also using Microsoft 365 Copilot to streamline repetitive and administrative tasks. Specifically, it helps with updating and managing the internal R&D governance manual, summarizing and updating documents, and creating PowerPoint presentations. This frees up time for strategic thinking and intensive problem-solving that leads to innovation.
According to Almirall, its goal is to build more efficient research processes that help accelerate drug development and, thus, promote true cures for people with skin diseases. "This project is fully aligned with our values," says Francesc Fernandez, senior director of R&D Data Science at Almirall. "By making better use of existing data, we can make faster decisions, reach the market sooner, and help patients sooner," he concludes.
EN 
PT 
