A machine learning model has been successfully deployed in a Databricks environment using MLflow. The model is now receiving a significant number of requests, and the data scientist needs to ensure high availability. What strategy should the data scientist employ to achieve this?