Deploying models
Once we generate a model that abides by our initial KPI requirements, we need to deploy it in the production environment. This could be something as simple as creating a list of neighborhoods and political issues to address in each neighborhood, or something as complex as shipping the model to thousands of machines to make real-time decisions about which advertisements to buy for a particular marketing campaign. Once deployed to production, it is important to keep on monitoring those KPIs to make sure we're still solving the problem we aimed at initially. Sometimes, the model could have negative effects due to a change in trends, and another model needs to be trained. For instance, listeners over time may lose interest in continually hearing the same music style and the process must start all over again.