CUSTOMER INTELLIGENCE CENTER


Challenges
The project is to replace old legacy system CIC 1.0 which is used by the car dealerships to send out promotions and offers to customers where the processing times have significantly increase due to the tremendous increase in data and to build a Machine Learning solution to identify what Marketing Contents to be pushed to which customers.

Solutions

  • Used capabilities of Amazon Web Services (AWS) to replace the CIC 1.0 legacy system and process on average 50 millions rows of production data.
  • Migrated MySQL Database to Amazon Aurora and due to the parallel capabilities of Apache Spark the processing times have been reduced significantly from 4 hours to 2.28 minutes.
  • Pick up the data and perform transformations and land it in Sandbox where it serves the purpose for Data science to identify feature set and develop forecast solution.
  • The Data Science Solution served the purpose to run targeted marketing campaign to the automobile users.

ROI

  • Decrease in the latency of the marketing channel data propagation by 85%.
  • Increase in the data throughput from tens of records to millions for real-time insights and analytics.
  • Integration with existing infrastructure to reduce the cost of ownership.