Customer success is everything for a company. It means staying ahead of customer needs, anticipating them in advance, improving their user experience, tracking and analysing their feedback, and swift resolution of their complaints. As a company scales up, anticipating customer complaints in advance means fewer customer complaints and better NPS (net promoter score). A large British bank set out to achieve this as part of their customer experience roadmap. Riddled with batch processing legacy systems, near real time complaint resolution was a tough case to crack for them.

One of our consultants tackled this problem head-on, creating a near real-time operational data store (ODS) that takes in customer interactions data (branch, call centre, POS etc.) to build a machine learning algorithm that can predict the propensity/likelihood of a customer complaint and push this insight real-time to service desk agents for proactive resolution. The team used Ab Initio for data integration and Mongo DB for data storage, leveraging Ab Initio’s continuous graphs ability, ingesting data from 20 operational databases, and transforming that in real-time to be fed to a MongoDB for consumption. The solution processed 25 million customer records daily in near real-time – voice, emails, push notifications, text messages, call centre, etc.

The implementation improved SLAs on complaints resolution, paving the way for many more customer personalization initiatives to enhance customer experience and drive NPS for the bank.