Operationalization of a Machine Learning solution for predicting drug demand
Client
A global Swiss pharmaceutical company
industry
Medicine and pharmacy
Competence center
Advanced Data Analytics
Technologies
R, GitLab, AWS, (EC2, S3, ECR), Docker, Tableau
Project description
The aim of the project was to create tools to predict the production needs of an influenza drug. The algorithm examines the occurrences of influenza in geographical terms and based on historical data, then the ML Ops infrastructure automates the process taking place entirely in the cloud environment.
The source code for R is in the GitLab repository. Pipeline collects historical data and converts it in the form of a model in the AWS cloud. The results are presented in real time in Tableau. Thanks to this, the client can estimate the amount of production needed, reduce its costs and predict the development of the disease.