Data instead of test tubes - KI-supported simulations for research and development

The research and development of new innovative products generally requires a high investment of resources and time - as is the case in the chemical and pharmaceutical industries. In order to make the development processes for these companies more efficient, the Berlin startup is developing the software solution Quantum on Demand, which enables them to integrate simulations into their research and development processes. In an interview, the three founders Dr. Marcel Quennet, Dr. Gunter Hermann and Dr. Vincent Pohl, who also participated in the 5-HT-Programm X-Linker discuss how their concept can optimize research in chemical and pharmaceutical companies.

What is QoD Technologies?

Marcel: We are developing a software solution for fully automated chemical prediction. These are based on state-of-the-art simulation methods and artificial intelligence and make it possible to predict the properties of materials and active pharmaceutical ingredients and parts of experiments on the computer. As a result, the synthesis of active ingredients and the performance of experiments in the laboratory can be partially dispensed with. This can result in significant time and cost savings for chemical and pharmaceutical companies in the research and development of novel chemicals, materials and drugs.

What problems does QoD Technologies solve in the development of new chemical products ?

Gunter: Since the development of new chemicals and active pharmaceutical ingredients is a very cost- and resource-intensive process, simulations are increasingly being used to shorten the development time of these products. However, only large corporations can afford the expensive software and infrastructure as well as the necessary experts. With QoD Technologies, we provide a complete package for simulations in a single software solution (Quantum on Demand) and with an attractive price/performance ratio, especially for SMEs .

Which concrete research projects can be facilitated with your solution ?

Vincent: One example is the development of a new organic solar cell. Here, the absorption properties, i.e. in which part of the sunlight the desired compound is absorbed, and the structural stability are important parameters. Using our software solution, researchers can identify the most promising compounds among the potential candidates and optimize them on the computer. This saves them many steps in the laboratory. For example, not all candidates have to be synthesized and characterized in the laboratory.

Gunter: In general, a large number of measurable chemical and physical properties can be predicted. These include structural, spectroscopic and thermodynamic properties as well as binding and adsorption energies. We are constantly working on expanding this spectrum.

What is so special about your technology ?

Marcel: Our goal is to make our data- and simulation-based predictions accessible to every researcher in the chemical and pharmaceutical industries. In order to keep the entry barrier as low as possible, three components are of great importance. First, we have automated the entire simulation workflow. This usually requires many small-step manual processes and a high degree of expertise in selecting the appropriate simulation method. On the other hand, we provide companies with the necessary software and IT infrastructure. The last component is the seamless integration of our solution into the workflows of the researchers. In addition to a user-friendly web application, we have developed an interface that makes it easy to integrate our solution into existing software, such as electronic laboratory journals.

Gunter: Another important component of our software solutions is the application of artificial intelligence for chemical and biochemical predictions. These are based on a large data set of experimental and simulation data. On this basis, we can accelerate and improve predictions for our customers.

How did QoD Technologies come about ?

Vincent: The idea was born during our promotions. We have found that simulations are an important tool for experimenters to gain deeper insights into their research. However, they often lacked the necessary expertise to carry out the multitude of complex sub-steps independently. We are therefore convinced that the experimenters could benefit from a software solution that performs these steps fully automatically. With the financial support of the Berlin Startup Scholarship, we began validating our business idea in September 2018. Since April, we have been supported by the EXIST startup scholarship.

How far are you currently in development ?

Marcel: In the meantime, we have developed a proof of concept for our software, which we are currently testing with our first pilot customers. In addition, we work with external partners to integrate our software into other software, such as laboratory journals. After the evaluation and further development of our software, we want to carry out a public beta test as soon as possible.

Where do you see each other in five years ?

Vincent: In five years' time, we will have expanded our product range to such an extent that simulations of highly complex compounds such as high-performance materials can be automated. Of course, these predictions based on intelligent data linking combined with artificial intelligence should be many times faster than is currently possible .

Together with eleven other startups, you took part in our X-Linker program in May 2019, where there were various mentoring sessions and where startups could network with investors and corporates. What did you take with you ?

Marcel: At that time, we were still in an early phase with QoD Technologies, especially in the second month of the EXIST founder scholarship. In addition to the pitching sessions, the mentoring sessions on topics such as business model, product and especially team were very valuable for us. We gained extensive insights into the various topics. The feedback from the mentors helped us to think through some details of our business model in more detail, but also confirmed us in many areas. That was especially helpful in our early phase. At Corporate Speeddating, it was interesting to see how large corporations rated our product. In addition, making contact with those responsible in the companies was very beneficial. The contacts made there will help us to initiate projects.