Datacraft Enables SBI MF Speech-Based IVR Deployment

One of the most fascinating things about India is its diverse culture. With cultural diversity, different languages and accents come into play. In order to cater to the citizens of this demographically vibrant country, SBI Mutual Fund sought to put in place a speech-enabled Interactive Voice Response (IVR) solution.

In its pursuit to make this initiative a success, it partnered with Datacraft to create a complete technology platform to ease all operations and install varied customer interfaces. For this, SBI Mutual Fund opted for Nuance Speech Recognition solution along with the Cisco IP IVR and both these were deployed using the expertise of Datacraft Professional Services.

Need for a speech-based solution

SBI Mutual Fund wanted to use speech recognition to provide Net Asset Values (NAVs) of its various Mutual Fund schemes to its customers via touch-tone customer services. SBI Mutual Fund has a long list of funds, which cannot be serviced by dual-tone multi-frequency (DTMF) IVR. Therefore, it required a speech system, which would act as a self-service platform and provide future flexibility.

After much brainstorming, the company zeroed in on an Interactive Voice Response with speech engine as it offered the feature of prompting the caller for the fund name. The system recognises the fund, retrieves the latest NAV value from the back-end database and finally plays out the NAV values to the caller. The system also allows callers to request for NAV information of up to five funds.

End to tiresome wait periods

Prior to launching the IVR solution, the customers had to go through a call centre and be addressed by the call centre personnel. Sunil Mangalore, CEO, Datacraft India, said, "This service was time consuming and very people dependent. The quality control and the level of service was very call centre person oriented. It was critical to address the customer immediately and also provide him/her with the information within a stipulated time. Privileged customers also needed to be treated with some priority and a sense of importance. The speech recognition-based IVR solution enables clients to speak to the IVR system to get the latest NAVs of various SBI MF schemes."

Major challenges during the design phase

The profound competition in India’s financial services sector and the high level of customer expectations in terms of service always pose challenges to providers. The deployment of the solution chosen by SBI MF was complex due to various reasons like non-standardisation of mutual fund names, which neecssiatted a lot of customisation. The other major challenge was that the SBI MF customer base spoke in various dialects.

Mangalore says, "The Datacraft team had to listen to hundreds of customer interactions to track the trends in speaker dialects and the manner of requesting NAVs of mutual fund schemes. The NAV requests were unique as customers would use the full name, short form or abbreviation of the fund name. Before the solution went live, the team conducted test trials for over 2000 calls so as to fine-tune the application to adapt to different dialects and make it accurate." In spite of these challenges, the project was completed in 90 days.


SBI MF has always given priority to customer service. This implementation has resulted in improved customer services and has also helped in providing NAVs of mutual funds instantly to callers.

The IVR solution has facilitated automated routine, quick response to frequent enquiries and has left agents with more time on hand to address complex queries. In terms of cost, the operator costs have reduced by 30 percent as per company claims. The implementation of this solution has also resulted in 20 percent more accuracy in the speech recognition application that will provide a concrete RoI in 15 months.

Find latest and upcoming tech gadgets online on Tech2 Gadgets. Get technology news, gadgets reviews & ratings. Popular gadgets including laptop, tablet and mobile specifications, features, prices, comparison.

Updated Date: Sep 09, 2009 12:00:04 IST