Data Reconciliation
Telecommunication
Charges for end customers were not calculated correctly.
Dashboard, ML model
INITIAL SITUATION
Our customer experienced issues in which users were occasionally overcharged or undercharged for their mobile plan.
This was due to inconsistencies between CRM and billing data.
SOLUTION
A dashboard was built in which inconsistencies could be easily visualized.
An underlying ML model was used to detect and prevent such inconsistencies.
BUSINESS VALUE
The revenue assurance department could detect and prevent over/undercharging.
In this way the customers’ satisfaction and the financial transparency increased.
FRAMEWORK & TOOLS
Pyspark, Sklearn, Oracle, Pandas, Tableau</span