Preparing for deployment in production: key considerations
In this section, we will look at a typical architecture deployment for intelligent applications. There are a lot of other aspects we need to keep in mind when we are moving to production. For simplicity, we will refer to the augmentation application we built in Chapters 9 and 10.
Let us look at all the tasks we did from loading data to reviewing the results:
- We loaded the data into a graph.
- The graph was enhanced with seasonal relationships. We augmented the graph using the augmentation application—articles as well as customer behavior aspects.
- We also utilized KNN similarity and community detection algorithms to enhance the graph and reviewed how this approach gives us better results.
In a production deployment, all of these aspects may need to be automated and deployed as individual applications. Let us take a brief look at all of these aspects that we need to take care of.
...
