Success Story
Automatic Categorization for 98.5% of Card Transactions
With millions of clients conducting millions of operations every day, the bank needed to automatically assign a unique category to every banking transaction (card as well as non-card).
2x
increase in categories assigned to card transactions
40%
increase in the accuracy of categorization for card operations
Challenge
With millions of clients conducting millions of operations every day, the bank needed to automatically assign a unique category to every banking transaction (card as well as non-card).
That’s why we developed an application that consolidates all types of categories into one structure, i.e., a single categorization tree.
- The application automatically categorizes 2.1 million banking transactions every day.
- Currently, 25 different Advanced Analytics and Data Science models are used in the bank to categorize primarily non-card transactions
Categorization contributes to a better understanding of financial behavior.
Impact
The banka (retail, marketing, operations)
- Obtains reports and aggregated data for the customer segments being monitored
- Better targets and personalizes communication with customers
- Uses categorization as a basis for banking advisors and automated advice services
The client
- Gains an accurate overview of their income and expenses
- Can independently create and manage their own categories
- Acquires a clear and unambiguous resource for adjusting their expenses and investment portfolio
- Increases their financial literacy and stability
