Chat­bots, AI in NBFCs

Find rapid tri­als by cus­tomers:

Banking Frontiers - - News - [email protected]­cal­in­fo­

The great Amer­i­can aca­demic Erik Bryn­jolf­s­son shared his thoughts on ar­ti­fi­cial in­tel­li­gence and its use­ful­ness in rou­tine work life. He said there are lots of ex­am­ples of rou­tine, mid­dleskilled jobs that in­volve rel­a­tively struc­tured tasks and those are the jobs that are be­ing elim­i­nated the fastest. Those kinds of jobs are eas­ier for our friends in the ar­ti­fi­cial in­tel­li­gence com­mu­nity to de­sign ro­bots to han­dle them. They could be soft­ware ro­bots; they could be phys­i­cal ro­bots.


Moti­lal Oswal Fi­nan­cial Ser­vices uses ar­ti­fi­cial in­tel­li­gence in mul­ti­ple places for analysing cus­tomer be­hav­iour in terms of cus­tomers’ needs and pref­er­ences re­flected by in­ter­act­ing with them us­ing dif­fer­ent chan­nels and touch­points. This helps the com­pany to de­sign the prod­ucts in a bet­ter man­ner, which are highly rel­e­vant to the au­di­ence. It also helps to cus­tom­ize the ad­vice given to the cus­tomers.

Arun Chaudhry, head - On­line Busi­ness & Prod­uct De­vel­op­ment at the com­pany, re­veals the de­tails about the use of AI in the or­gan­i­sa­tion: “We are ex­ten­sively us­ing AI in build­ing au­to­mated ad­vi­sory prod­ucts wherein the en­tire process from risk pro­fil­ing, in­vest­ment sug­ges­tion to track­ing is au­to­mated with lim­ited hu­man in­ter­ven­tion.”

Ra­jiv Rai, chief dig­i­tal of­fi­cer at Edel­weiss Fi­nan­cial Ser­vices says AI is see­ing mul­ti­ple de­ploy­ments and use cases. “It is be­ing used for read­ing cus­tomer emails, cat­e­goris­ing them and auto re­spond­ing (to some cat­e­gory). We are also us­ing AI to en­able credit de­ci­sion­ing on-the-fly by be­ing able to read and an­a­lyse var­i­ous doc­u­ments pro­vided by the cus­tomers like bank state­ment, P&L state­ment, salary slips, etc,” he adds.


Cus­tomer ser­vice, sales and hu­man re­sources are the top 3 ar­eas in which NBFCs uses chat­bot plat­forms. The plat­form is use­ful for prod­uct stream, in­fras­truc­ture, tech­nol­ogy, in­sur­ance, lend­ing, of­fice man­age­ment, cus­tomer ac­count han­dling, etc. Ac­cord­ing to Sudip Ban­er­jee, CTO at Re­liance Cap­i­tal, the com­pany mainly uses the chat­bot plat­form for pro­vid­ing ef­fec­tive cus­tomer ser­vice, the queries han­dled by call cen­ter is trans­ferred to chat­bot as it does not re­quire hu­man ef­forts to do the job. “Un­for­tu­nately, voice recog­ni­tion is not pos­si­ble in In­dia, as cus­tomers ask ques­tions in ‘hinglish’,” he says.

Edel­weiss Fi­nan­cial Ser­vices started us­ing chat­bots by de­ploy­ing them for its em­ploy­ees. This brought 2 ad­van­tages. It is able to fine­tune the chat­bot strat­egy by de­ploy­ing it on a con­trolled set of users. Says Ra­jiv Rai: “Chat­bot ser­vice has be­come an in­te­gral part of our ser­vice, of­fer­ing it both for our em­ploy­ees as well as cus­tomers.”

Moti­lal Oswal Fi­nan­cial Ser­vices uses chat­bots for solving all kinds of rou­tine re­quests - for ledger state­ments, con­tract notes etc. More than 50 such re­quests are ad­dressed di­rectly by the chat­bot. The core idea is to cir­cum­vent a typ­i­cal IVR and get straight to the point by giv­ing the re­quired re­sponses. “We can en­hance the cus­tomer ex­pe­ri­ence by pro­vid­ing in­stant and ac­cu­rate re­sponses to the cus­tomer queries which would have oth­er­wise in­volved

a rather long and man­ual process”, says Arun Chaudhry.

He adds on the cus­tomer re­sponse re­ceived from the chat­bot plat­form: “Our chat­bot MO Ge­nie is a 100% in­ter­nally de­vel­oped self-learn­ing sys­tem, which to­day is ca­pa­ble of han­dling 50 dif­fer­ent types of client queries. Till date MO Ge­nie has been able to ad­dress more than 1 mil­lion queries, which were raised by more than 70,000 unique users; it cov­ered more than 70% of the over­all cus­tomer queries. We have no­ticed a more than 30% in­crease in the count of queries han­dled by Ge­nie in the last 6 months, which now pro­cesses more than a lakh re­quests each month.”


The chat­bot plat­form is mainly in used in English lan­guage by the NBFCs in In­dia. Some of the com­pa­nies now in­tend to ex­pand the scope of the chat­bots, in­tro­duc­ing other In­dian lan­guages as well. Edel­weiss Fi­nan­cial Ser­vices uses nat­u­ral lan­guage pro­cess­ing and neu­ral net­work­based al­go­rithms to de­velop these ser­vices.

“We have not seen any par­tic­u­lar us­age bias. Peo­ple are us­ing the ser­vice from both mo­bile phones and com­put­ers. What the cus­tomers love about these ser­vices is that there is no lead time in re­sponse and these ques­tions can be an­swered at any time of the day or night. We use a panel of ser­vice providers from Pune, Hy­der­abad and Ban­ga­lore for the chat­bot and AI ser­vices. We also have a ded­i­cated team in-house whose ef­forts are sup­ple­mented by these part­ners”, says Ra­jiv Rai.

Re­liance Cap­i­tal mainly uses tech­nol­ogy of the star­tups and fin­tech com­pa­nies. The com­pany also uses Mi­crosoft, IBM Wat­son and Ama­zon tech­nol­ogy for AI and chat­bot.


There will be 3 kinds of bots pri­mar­ily text-based, speech-based and video-based. Al­most all the ma­jor NBFCs are now likely to cre­ate a con­ver­sa­tional app, where even voice can be used for au­then­ti­ca­tion. Chat­bots and con­ver­sa­tional lend­ing are the way for­ward. These will make whole process eas­ier and fric­tion­less for the NBFCs and their cus­tomers.

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.