A fitting deployment strategy
You have realized that your current model is severely degraded, so you have collected a new batch of training data and trained the new model.
Your model pipeline hasn't changed in any way. You have just re-trained it on new data. Therefore, you are confident it is safe to immediately start serving this new, better-performing model to all your users.
Which deployment strategy makes the most sense in this case?
This exercise is part of the course
MLOps Deployment and Life Cycling
Hands-on interactive exercise
Turn theory into action with one of our interactive exercises
