YES BANK is India's fourth largest private sector bank. With over 1,100 branches serving 2m+ customers in India, YES BANK is on a mission to become a “Technology Company in the Business of Banking.”
YES BANK has many systems generating structured and unstructured data, from hundreds of applications, click streams, logs, emails, text messages, push notifications and other sources. The bank’s federated data marts gave each business only enough data to substantiate its own business, without a 360 degree customer view. With the bank’s focus on delivering customized services across its entire value chain it was imperative to get a unified view and maximize its ‘data banks’.
“We aim to completely personalize the customer’s experience using data services,” said Anup Purohit, Group President and CIO at YES BANK. “Our goal is not to create customer segments - but to treat every customer as a segment.”
YES BANK needed a solution framework to provide speed, agility, flexibility, and storage capacity to process unstructured data and run real-time analytics while heightening data security. Essential to this work was implementing a platform that could meet the high governance standards and stringent data security regulations of the financial services industry.
YES BANK partnered with Cloudera to build a ‘unified on-premise data management platform’. The Cloudera Shared Data Experience (SDX) technologies built into the platform mitigate any anxieties over compliance and ensuring secure data access across many users.
Simple, secure, and efficient data storage and management allows the team to devote more time to analytics. Employees across the company can run their own analytic reports and they can even set up automated predefined reports for key stakeholders. With self-service access to real-time data, businesses now have a comprehensive view of their customers, which directly translates knowledge into improved decision-making and performance.
YES BANK leverages the benefits of machine learning through programming languages like python and R to consolidate all customer data, personalize services thereby transferring the benefits of analytics directly to customers. The bank built a neural-network-based transaction purpose model, which looks at roughly 1,000 features and classifies the true purpose of every single customer transaction. The results have been a game-changing digital experience with proactive customer service.
Anup further explains, “For example, if we know that your wire transfer is actually a rent payment, or a school fee payment, we can use this data to build a reminder experience, which would include details like the remitter, the beneficiary, the amount, the duration, etc.”
This new information and insights now serve as a bedrock for the bank to build further machine learning models and continue to improve the customer experience. The Bank is building an additional neural network model to more granularly classify and identify customer demographics. The objective is not to fit customers into traditional demographic segments, but rather to create a personalized segment for each individual customer.
YES BANK is continually exploring ways to use data as a key driver to improve customer acquisition and overall customer experience. “Data is vital not just to ‘target’ customers but also to add value to the experience we give them,” says Purohit. For example, the bank is building intelligent and proactive chatbots that will know not only what clients need, but when they need it. “Reaching out to the customer at the right time is of critical importance and we have a lot of emphasis on running this in real time.”
With a unified data platform built on Cloudera, YES BANK now develops be-spoke and inventive solutions with quicker turn-around times for product launches and updates. Empowered by the accessibility and agency over its data, YES BANK has brought a previously out-sourced customer loyalty program in-house, which has saved them a quarter of a million dollars every year.
“We are increasingly moving towards a culture where all our decisions are based on data and not ‘gut feel’.”