While high-tech industries have rapidly adopted AI and ML, the biopharmaceutical industry has also shown interest in modern technological solutions. Our client is a German company that provides software solutions for the biopharma industry in predictive biomanufacturing.
The client leverages Digital Twins to model biochemical processes using predictive models. It helps to analyze metabolic processes using real-time customer data and, as a result, speeds up the manufacturing of medical drugs and their time-to-market. The project aimed to optimize the code in several modules and assist with the research and development phase.
At the start of the cooperation, the client requested two Python teams.
- One team was working with the client's development team on tweaking the code for modules and completing POC development to increase the sales conversion of the product. The output of this cooperation has been increased speed of the solution and more stable and accurate code.
- Another team was responsible for researching and optimizing open-source modules that could be used in the modulation process. If an original module had 30 parameters, the Elinext team reduced this number to five to achieve better speed of bio-data processing. Then, with the AI algorithm we wrote, the system would expand the number of parameters to output the information to other modules or display them on graphs.
Elinext researched ways to optimize the application with the affiliate Amazon Cloud solution. The optimization made it possible for instances to work faster together.
The cooperation between the client and Elinext lasted for three months. Starting with a smooth integration of our team with the client's, we optimized and increased the speed of the code and the application itself. We also explored the possibility of speeding up calculations on the server by introducing an information coder/encoder.