Operationalizing ML
As discussed in earlier chapters, you can enjoy the full benefits of ML in your business if your models get deployed and used in the production environment. Operationalization is more than just deploying the ML model. There are also other things that need to be addressed to have successful ML-enabled applications in production. Let's get into it.
Setting the business expectations
It is extremely important to ensure that the business stakeholders understand the risk of making business decisions using the ML model's predictions. You do not want to be in a situation where your organization fails because of ML. Zillow, a real estate company that invested a lot in ML with their product Zestimate, lost 500 million dollars due to incorrect price estimates of real properties. They ended up buying properties at prices set by their ML model that they eventually ended up selling for a much lower price.
ML models are not perfect; they make mistakes. The business...
