💰Fraud identification

Use case: This tool can identify different types of credit card fraud based on account activity and transactional behaviour.

End-user: Fraud case handlers.

Solution: a Rainbird agent was deployed that case handlers could use to ask about an account or transaction. The Rainbird Knowledge Map was connected to a datasource to retrieve the relevant account or transaction data, resulting in the case handler not being asked to provide the data themselves. For this datasource the Firebase API was used with the data being stored in Google Sheets, but many other datasources are supported.

Expertise: Replicates the expertise and decision making ability of fraud case handlers at financial companies.

Data: Data is stored in the Knowledge Map, with case-specific data being stored and retrieved from Google Sheets at runtime.

Last updated