send a json request to the api once a user executed a target behaviour.
bundle the data into users’ profiles and model it with a joint statistical modelling strategy.
feed the algorithm with the training data and help the model learning.
the model generates a score which evaluates the possibility of users’ fraudulent motive.
label the user as “fraud” or “not fraud”.
use for loan application inspection and post-loan management.
reduce the non-performing rate of loans. prevent fake transaction.
provide the latest model deployment and tailor-made modelling services.
help lenders assess the borrower’s credit quickly and accurately.
it has been tailor-made for four countries, including indonesia, the philippines, vietnam and india.
start the conversation with our team. we’re always happy to help.