A key element to the result accuracy achieved using Zensed anti-fraud Machine Learning is its trained models and its isolated architecture. The Zensed models are strong enough to evolve and adapt their machine learning whenever there is a new fraud trend emerging or more information is available to a given transaction.
Using Zensed is friction free. Merchants do not have to alter their systems or business process in any way to use our technology. Integration is straightforward and Zensed will provide clients with technical consultancy during the set up process.
Our clients pass transaction information to Zensed using our REST API. The Zensed system uses this information and automatically cross references this data with fraud patterns and trends in the Zensed proprietary modelling platform.
These pattern and trend models have been built and refined after analysing millions of client transactions and the Zensed machine learning platform is designed to continually improve results; the more data we receive the better Zensed results get.
The Zensed system rapidly analyses all the cross referenced patterns in the data related to the client transaction and provides a score presented as a risk percentage regarding how likely it is that the transaction will need to be refunded or will cause a chargeback.
Our clients have their own portal access to the Zensed back office system where they can react to all their transaction results. Where a transaction has been identified as potentially fraudulent this is marked with a certain fraud risk percentage or in a certain risk group. The client then decides if they wish to block a transaction or perhaps add extra validation checks where a transaction is a border line case.
Zensed presents the predicted outcome to the client with a score based upon likely fraud and risk of chargeback. The client can act on the Zensed result as they choose although we do provide our experience and findings to clients as part of the business relationship as and when required.