The cost of developing new drugs is now widely acknowledged by industry leaders as prohibitive, with some estimates now in excess of $ 5B per drug. It is imperative that new approaches are developed to mitigate the cost and time required to bring new drugs to market. To help achieve this, we contend that new techniques in machine learning must be brought to bear on the drug development process. The quantity of data generated nowadays is amenable to machine learning where, by analysing the data using special algorithm, we can begin to predict if a compound may be efficacious and, in addition, make informed decisions to improve upon it. Machine learning as proven, in different application areas, to be an effective approach to perform predictions from data and to improve certain tasks. We plan to use this approach to help in building better drugs by improving the process of finding molecular structure using the building blocks of proteins, the amino acids, which could one day serve as models to create drugs against microbes. The process could be extended to other fields (cancerology, neurology, infectious and immunological diseases, among many others) since the methods developed can be adapted easily. To bring this project to fruition, we have formed a team with diversified expertise (consisting of microbiologists, chemists, computer scientists, and mathematicians) with the hope that it will revolutionize the way drugs are made, saving both time and money whilst creating better and safer drugs.