In our previous coverage of Yes Bankâ€™s Datathon, we came across a very unique proposition. The 8 teams guided by YES Bankâ€™s years of market expertise were able to bring forth some prodigious and unforeseen solutions, rather â€˜modulesâ€™. In few of the illustrations in our prior story, we introduced you the modules that were designed and intended to solve some of the most primitive and time-consuming challenges. This installment of the event comes after the winners of the Datathon are announced. In our rendezvous with the eventâ€™s mega minds, the spotlight was thrown on the two highly effective models – the â€˜Financial Transaction Community Detectionâ€™, and the â€˜Next Best Actionâ€™.
Next Best Action
Datathon was conceived with the aim to outsource not only the ideas and concept but to get an end to end data module state. Including the two modules, all the other modules are being worked on to integrate with government banking, and mobile wallet etc. However, the most profound and seldom discussed requirement is the â€˜next best actionâ€™ feature. Given a customerâ€™s financial indulgence, timely credit & debit history, perks available on his/her credit card and several other aspects, the module is designed to suggest the next best action to the customer. One part of the module is being developed by the data scientists that have internally observed the development of the idea, but a larger part of it was created as part of Datathon, wherein they looked at the product history, transaction history of the customers and look at the funding throughout.Â For an instance, if a certain amount is being transferred through an investment platform, the module will classify it into share trading and do the relevant prediction.Â Presently the winning team that engineered this module is working with the internal team, doing fine-tuning and testing with internal customers. The module is also developed as APIs for YES Bankâ€™s mobile banking model. The module is, in fact, being built to blur all the lines between platforms and maintain one single reference. Irrespective of a customer looking into products on different platforms, interests shown or neglected, cards blocked or service requests registered, the next best action module will be able to take all this into account and suggest the best prescription for the financial health. The time to market for this module is shrinking as the team is stepping up their efforts with each passing day.
Understanding your Network- Financial Transaction Community Detection
The banking industry can vouch for the second model that shows promise to further enable banks to reach out to their customers and potential customers effectively. The â€˜Financial Transaction Communityâ€™ module is ideated by the students from VIT Vellore. Based on a customerâ€™s financial or transactional contacts the module builds a profile to understand his/her spending techniques, and in turn, identify the potential customers. With respect to customerâ€™s financial relationship with his/her peers or acquaintances, the module creates maps or profiles that understands the whole ecosystem surrounding the customer. For instance, if there is a service that person A in his/her network is not using, it can be cross-sold or off sold to person B. So the idea is to sell more. This is not for walk-in progress but is being looked upon to develop many packets of it. The model is scheduled to get released by sometime around June next year. Being a customer facing module, the results will be provided to the relationship manager to make their life slightly easier. Once this model goes live, YES Bank intends to take on a much more challenging customer base of the corporate world.
As mentioned earlier, there are 8 modules in total that are in production pipeline , but the aforementioned two are expected to make the maximum impact on the banking industry. YES Bank has been building applications around the models. Given the humongous set of real-time data that YES Bank has carried all along, it would be a mere surprise if the banking industry takes a sharp take off to the most efficient banking system seen in decades.