Targeted protein degradation has been one of the merging technologies to solve uncurable diseases like Alzheimers, Huntingtons etc. The problem is that the process of this biotechnology in labs is very long and expensive. Automating this insilico (with computers) and machine learning is a challenge we at Celeris Therapeutics are tackling. Doing research in chemistry, biology and machine learning is how we are solving this problem. Our main component is the emerging field of ML called geometric deep learning. Graphs are native way to represent atom and molekular interactions, but given that the space of possible compounds is huge this becomes a big engineering problem. And thats what we are trying to unite at Celeris, biotechnology, engineering and machine learning.