There has been every increasing interest of AI in banking and other facets of the financial sector. Several media talks and market research around the emergence of AI in the banking industry range from the topics of wealth management , digitizing paper documents , insurance claims , credit scoring and risk management . Among all the above stated growth areas of AI, automating risk management processes are of highest interest due to best ROIs and shortest time to market.
Chatbots on the other hand are now considered to be over hyped due to the limited efficacy of natural language processing models and the lowest ROIs.
In order to ensure Digitisation, banks have facilitated online transactions. Although it has provided a great deal of ease for their customers, it exposes several online threats such as identify theft. I.e., how to ensure that the person making the online transaction is the original account holder himself.
We have developed scientific methods to help banks catch and block any such fraudulent transactions. Our time series analysis methods learn from the transaction history of the customers and sketch transaction boundaries of individual customers. This helps to identify transaction anomalies and leads to fraud detection. The anomalies generate alarms and at the same time block the suspicious transactions which can then be manually analysed and intervened.
