AI as an assistant in research and development 

Artificial intelligence (AI) can help to develop medicines more efficiently for patients. Did you know that Boehringer Ingelheim is one of the three percent pioneers in AI use? Get to know two of our AI tools here. 

According to a YouGov study1, only three percent of companies in Germany use their own AI tools. In contrast, 33 percent of the industry neither use nor plan to use AI. This is even though AI is considered a key technology for the transformation of industry which could make Germany competitive again. 

What is artificial intelligence and how does it work in the world of work? 

In our case, AI refers to the development of computer programs. They can perform tasks in the workplace. They can identify patterns, draw conclusions, and solve problems. At the same time, these systems can learn to improve their performance over time. There are different approaches and techniques in AI. For example, machine learning, neural networks, natural language processing and expert systems are all part of this. 

Concrete AI applications in research and development 

Two clicks for a summary of a text versus endless hours of searching through literature. Three clicks on the screen for predicting compound properties versus days of synthesis in the laboratory. Both scenarios show the strengths of AI: with the right tools and the right application, it saves time and resources. 

AI increases the probability of success in research and development. It thus brings us faster to medicines that can improve human and animal health. Our employees in many departments are already working with AI tools. They are in regular contact with tech companies and other AI users from the industry. The aim is to discuss the role of AI for the working world of tomorrow and to exchange experiences and opportunities. For example, as part of the "AI Future Tech" project of the eco Association of the Internet Industry (eco Verband der Internetwirtschaft), in cooperation with Microsoft Germany2.  Boehringer presented iQNow and the molecule designer ADAM as a best practice – but what distinguishes these two tools? 

AI Tool iQNow - our own ChatGPT 

Michael Schorpp
Michael Schorpp, Head of iQNow & AI

Michael Schorpp, Head of iQNow & AI 

As an in-house ChatGPT, "iQNow" is more than just a knowledge management system - it is an assistant for our employees. With iQNow, the quality of work processes can be improved and repetitive work steps can be taken over by AI. 

The platform was developed over four years ago to collect all the company's knowledge. Thanks to the Large Language Model and ChatGPT, knowledge is now even easier to access. For our scientists, this means that they no longer have to spend hours reading literature to know the most important results from research.  

 

Instead, you can ask iQNow about the most important statements of a text or have short summaries written for you. With the right "prompts", in other words the tasks that you enter into ChatGPT, work steps in other areas can also be made easier. The data quality and security at iQNow is ensured by its ability to send documents for analysis from iQNow to iQGPT. 

Michael Schorpp, Head of iQNow and AI, sums up the advantages of iQNow: "It allows the scientist to extract knowledge very quickly, from an incredible number of documents, also from documents that are very rich in content. The purpose of the whole thing is to streamline our processes and get the drugs to the patient faster." 

AI Assistant ADAM - A virtual co-pilot for our research on molecules 
Nils Weskamp
Nils Weskamp, Head of Data Science in Computational Chemistry

While iQNow can be used by everyone, ADAM was developed for our researchers in medicinal chemistry. In chemical research, large amounts of data have mainly been evaluated manually.  

Thanks to machine learning and AI, data analysis is now automated. ADAM – short for "Advanced Design Assistant for Molecules" – supports the search for new molecules. It uses our substance library and published research literature. 

"Such methods help us to be faster and increase our chances of success," explains Nils Weskamp, Head of Data Science in Computational Chemistry.  

ADAM helps our molecule designers to identify new substances quickly. After a thorough evaluation, our colleagues then decide which candidates will be sent to the laboratory. 

Video on artificial intelligence at Boehringer Ingelheim 

Immerse yourself in our AI world: 

Sources
  1. https://www.eco.de/presse/eco-yougov-umfrage-ein-drittel-der-deutschen-unternehmen-verschlaeft-ki-revolution/

  2. https://www.eco.de/news/eco-podcast-zu-ki-future-tech/