Home > AI Solutions > Artificial Intelligence > White Papers > Using Machine Learning to Recommend Service Product Offerings > Architecture
The ML pipeline is configured to generate recommendations based on the following steps:
1. Retrieval of training data from the data lake.
2. Application of data pre-processing on the training data.
3. Training of multiple models on the data using the following algorithms:
4. Merging of results from K-Means clustering and MBA to generate recommendations for the customer.
5. Saving of recommendations on the data lake.
6. Pipeline triggering at scheduled intervals to check for upstream changes or feedback and to re-compute and initiate revised customer recommendations.
The recommendations that are produced are served by a Rest API endpoint (Figure 4). Rest API is integrated with the business platform to send the recommendations in the form of requests and responses to the customer.