Success case

Model for detecting fraud

in international transfers

The challenge

Banks face attempts of fraud, including identity theft. The Criminal Offenses team is a department responsible for detecting various fraud attempts and blocking operations identified as fraudulent.


The goal is to improve and expedite the detection of fraud related to international transactions. The model prioritizes transfers with a higher probability of being fraudulent and enables the security team to detect them more efficiently.

How have we done it?

  • Machine learning
  • Gradient boosting (XGBoost)
  • Word2Vec

Results provided

The solution helps the Criminal Offenses team prioritize fraud cases for analysis. The system enables banks to be more efficient and quicker in fraud detection, facilitating the recovery of defrauded funds and improving customer satisfaction.