Yes Bank creates ‘Yes Robot’ using Microsoft Cognitive Services

Lindsay James
Lindsay James
By Lindsay James on 27 September 2019
Yes Bank creates ‘Yes Robot’ using Microsoft Cognitive Services

India’s Yes Bank has created the ‘Yes Robot’ – an artificial intelligence (AI) powered 24/7 personal banking assistant which uses Microsoft Cognitive Services.

Yes Robot helps customers with more than 25 types of banking transactions. It currently handles more than half a million interactions every month, with over 90% accuracy.

It also checks eligibility and helps customers and prospects apply for over 65 banking products, including loans, with real-time assistance. Not only can customers chat with the bot to find answers to queries, they can also perform several financial transactions, like bill payments, recharges and money transfers.

The bot has a repository of industry-related terms, phrases and business intelligence that helps it to learn common customer queries over time, enhancing output accuracy. The bank is planning to extend the chatbot capabilities for its partners―paving the way for impactful customer service.

“We were focused on building a holistic AI ecosystem equipped with natural language processing (NLP) components, machine learning models, and other AI services,” said Anup Purohit, chief information officer at Yes bank in a Microsoft case study. “Leveraging Microsoft Cognitive Services, we are now capable of taking the conventional user experience to the next level.”

Ritesh Pai, chief digital officer at Yes Bank said the increasing use of Yes Robot has taken a significant load from customer service teams and resulted in improved levels of customer satisfaction. “We have recorded a 30% quarter on quarter growth in transactions volume, and we expect accelerated growth in the future as we continue to add more banking services to the bot,” he said. 

Microsoft Cognitive Services have enhanced the stability and capabilities of the YES ROBOT platform, which includes efficiency in handling user query variations, accuracy in responding to user queries as well as user goal completion rate for transactions, thereby helping accelerate the speed of adding new features and time-to-market.  These features have also ensured that the platform is future-ready for integration with other cognitive tools and engines in a seamless plug and play manner.

Currently, the bank is involved in assimilating user-suggested feedback in YES Robot into a learning loop. Going forward, the bank plans to integrate other features, such as voice-based commands, which will be interfaced with a data lake comprising 360-degree customer information to further enhance user experience.

Number of views (1458)/Comments (-)

Comments are only visible to subscribers.