With around 2mln active shoppers and more than 90mln transactions per year, Tinka’s challenge is to process as many legitimate payments as possible without causing any user friction. In order to achieve that goal, they need help leveraging machine learning and AI capabilities to better know their customers, reduce risk, and ensure compliance with strict financial and regulatory standards.
All of this needs to be achieved while keeping with the client’s values of transparency and honesty. To be able to quickly adjust to changes in fraud and/or pay behavior, the models need to be easily retrainable. Moreover, model explainability is also an important factor for reducing bias and improving the prediction process.