Machine Learning Model Deployment -- A Simplistic Checklist

There are many things that can go wrong when moving your machine learning model from a research environment to a production environment. These checks are better done in sequence to confirm issue from the previous scenario doesn't carry over to the next step These issues can be handled by covering these scenarios in code. If model predictions don't match or partially match: These are some of the most frequent scenarios practically observed and often overlooked by data scientists and machine learning engineers while developing and deploying models to production. Deploying and maintaining ML models are as hard (if not harder) than developing them. Hope this quick article helped you avoid common pitfalls in your workplace.