A basic analytics journey starts with reporting the inquiries right and transferring them to more advanced machine learning areas. “Our analytics journey started with traditional technologies, where we had data-warehouses of the world. We were having difficulties in providing analytical capabilities at scale. The bank was not able to use the usual data management systems to process the information to a quantum which was required to handle the number of users,” says Rajat Kanwar Gupta, President and Head ML/Analytics Solutions and Development CoE, YES Bank.
The bank was also struggling with selectively identifying the data that needed to be analysed. Consequently, Gupta started to look for a modern data platform, where they had multiple choices like a traditional data warehouse or an open source ecosystem approach. The bank further conducted POCs and finalised on an open source big data analytics platform and Cloudera became its partner.
The bank abstracted a lot of complexity on the management of the entire platform and needed a secure and maintainable cluster as it is critical for banks. Cloudera empowers YES Bank with the real-time processed data on the large scale. Gupta further states that the bank has built a unified and scalable analytics platform which is running on machine learning and also driving its user inquiries to get customer insights, where all the data is processes from multiple sources at different frequencies.
The improvement has been significant in terms of time saving; today the bank’s reports get processed within minutes, sometimes even in seconds vis-à-vis 24 hour plus timelines previously.
Gupta further describes, “Our data lake is constantly growing with a new richness of data and now we have started to utilise this data directly for the customers. Moreover, the new models are directly been introduced in our digital channels so there is no gap in delivery and generation.”
Despite having the new customer acquisitions opportunities, since the data was enormous, the bank faced difficulty in crunching the data within available resources and time. Now with Cloudera’s platform in place, it gives them the space of being selective about the data. “The bank is also creating next generation recommendation engines, which are personalised in nature; hence, we generate enough events so that we can proactively serve the customers,” he adds.
In order to improve the services and create a deeper relationship with the existing customers, the bank’s omnichannel campaign platform is supported by Cloudera’s platform. YES Bank is partnering with the omnichannel solutions which will empower the campaigns, traditional SMS email servers and other digital channels.
“The team is the biggest contributor in the entire journey for whatever we think, imagine and build. In such scenarios, we usually have a data sciences team and data engineering team who are always working in complete collaboration with each other. We have built a full stack team, wherein the data sciences guys can also do the data engineering job,” he adds.