Problem: Whether in the pharmaceutical or perfume industries, production or R&D they all face the same problem every day: the identification of unknown substances. The associated question “Which molecules are in my sample?” spans all areas of chemistry. Finding an answer typically requires an expert hours to weeks and costs around 1,000 € to 10,000 € per sample.
Solution: During a 3-year interdisciplinary research project the AI-Software METIS was developed. As powerful machine learning tool, METIS can generate accurate suggestions for unknowns, solely from the acquired data. With these, even non-experts require only minutes for identification. METIS comes as seamlessly integrated software-hardware solution.
USP: While databases (the main competitor) contain only a tiny fraction of known molecules (0.01%), METIS can predict correct structures even for truly unknown molecules. It can analyze data while it is being generated and is therefore the only tool that allows full process automation. As AI METIS also learns from every piece of data thereby smarter every day.
Markets: Analytical Chemistry, Industrial R&D Labs, QC