AI in HR man­age­ment ...............................

AI is gain­ing an all-per­vad­ing role. HR func­tions are in­creas­ingly car­ried out us­ing AI and the whole con­cept is get­ting per­fected:

Banking Frontiers - - Contents - mo­[email protected]­ingfron­

Un­til re­cently, the pri­mary use of tech­nol­ogy in HR man­age­ment was to im­prove ef­fi­ciency and de­rive cost sav­ings by au­tomat­ing repet­i­tive tasks. This has changed and smarter tech­nolo­gies are help­ing HR teams to solve crit­i­cal busi­ness chal­lenges, drive ex­po­nen­tial per­for­mance im­prove­ments and even im­pact larger busi­ness out­comes and prof­itabil­ity. Here, ar­ti­fi­cial in­tel­li­gence is cre­at­ing a tran­si­tion for HR from ad­min­is­tra­tive to strate­gic to mis­sion crit­i­cal. It is a given fact that the next com­pet­i­tive fron­tier for busi­nesses is em­ployee ex­pe­ri­ence and the fu­ture of HR will be cen­tered on this as­pect and per­son­al­iz­ing en­gage­ment.

Tech­nol­ogy has helped to pro­vide real-time HR data avail­able for busi­ness de­ci­sions, yet man­ual meth­ods still are be­ing used to get the in­sights from this data. This nat­u­rally cre­ates a bot­tle­neck. Here is where AI is poised to step in and ex­tract in­sights from data and de­liver rec­om­men­da­tions in real time. It can also elim­i­nate com­mon hu­man bi­ases and i ncon­sis­ten­cies i n sen­si­tive de­ci­sion-mak­ing. De­ci­sions pow­ered by AI have the po­ten­tial to be faster at scale and more data-in­formed, as well as more con­sis­tent and un­bi­ased.

AI in HR has lim­it­less op­por­tu­ni­ties. It has proven roles in at least 5 ar­eas: in re­cruit­ment, in HCM, in em­ployee en­gage­ment, in em­ployee ben­e­fits and in learn­ing and devel­op­ment.

Kishore Po­duri, ED & head, Hu­man Re­sources, DBS Bank In­dia, says to­day AI is ev­ery­where, and HR is pretty much in the thick of things. “You name it, from chat­bots, RPAs, an­a­lyt­ics to au­to­ma­tion, and HR ser­vices is suc­cess­fully uti­liz­ing the full spec­trum of avail­able tech­nol­ogy,” says he.


At DBS there are chat­bots car­ry­ing out var­i­ous in­ter­ac­tions with in­ter­nal as well as ex­ter­nal stake­hold­ers and ded­i­cated. AI al­go­rithms are com­mu­ni­cat­ing with can­di­dates/em­ploy­ees and ad­dress­ing their first level queries at the time of in­ter­view or/ and post join­ing. “Our re­cruiters have ac­cess to the lat­est tools that an­a­lyze re­sumes and com­pare them with ex­ist­ing em­ployee data to iden­tify suit­able can­di­dates who match the job de­scrip­tion and are suit­able for the or­ga­ni­za­tion in the long run. Au­to­ma­tion in pay­roll and claims pro­cess­ing is al­ready in full ef­fect at DBS and en­sures smooth pro­cess­ing. How­ever, the im­pact of AI in this space could be taken a step fur­ther in de­tect­ing frauds in claims, max­i­miz­ing uti­liza­tion of ben­e­fits, and pro­vid­ing in­di­vid­u­als with cus­tom­ized pack­ages as per their needs,” says he.

He cites the in­stance of cloud-based learn­ing man­age­ment sys­tems, which use AI to make per­son­al­ized cour­ses, or cer­ti­fi­ca­tion rec­om­men­da­tions, help em­ploy­ees col­lab­o­rate, cre­ate com­mu­ni­ties of in­ter­est, and en­gage in mo­bile ed­u­ca­tion any­time, any­where.

He, how­ever, is cau­tious: “As op­ti­mistic as I am, of the use of AI to en­hance cer­tain

ex­pe­ri­ences, I am also a firm be­liever in the fact that AI could never re­place the hu­man touch, which is of the essence when it comes to be­ing an HR pro­fes­sional.”


He ex­plains: AI could never do a ‘tough talk’ with an un­der­per­form­ing em­ployee un­happy with his or her assess­ment or man­age the sim­plest case of griev­ances re­dres­sal be­tween two in­di­vid­u­als. It could never do an emo­tional anal­ogy of an in­di­vid­ual and de­ter­mine whether they are in align­ment with the cul­ture of the com­pany.

“Ex­pe­ri­ences and jour­neys,’ he says, “are of the essence to us at DBS. Our peo­ple must de­velop a di­rect con­nec­tion with the ethos of the or­ga­ni­za­tion. This is again some­thing we could not ex­pect AI to re­place a hu­man be­ing at. I do not think we are there yet where the hu­man mind could be com­pletely re­placed by AI.”


Dr Vishalli Don­grie, part­ner and head, Peo­ple and Change, KPMG in In­dia, be­lieves there are con­sid­er­able tasks that are repet­i­tive, low-value in most func­tions in­clud­ing hu­man re­sources and AI presents an op­por­tu­nity to au­to­mate these and help free up time for more strate­gic work. “For a mean­ing­ful im­pact, the area of fo­cus for HR to­day is Dig­i­tal Labour,” says she.

“The aug­men­ta­tion and au­to­ma­tion of hu­man labour is in­creas­ingly find­ing its way across HR pro­cesses (ex. re­cruit­ment, em­ployee on­board­ing, per for­mance man­age­ment etc) as well with busi­ness unit em­ploy­ees. Most com­pa­nies lever­ag­ing these tech­nolo­gies to­day are uti­liz­ing co-bots, ie. bots that work along­side hu­mans, and HR is get­ting in­volved in train­ing em­ploy­ees to work with such co-bots. The other area of fo­cus for HR to­day is how to man­age the dig­i­tal na­tive tal­ent that is com­ing in to­day and how to build the right mind­set to lever­age their skills across the or­ga­ni­za­tion,” says she.

Ac­cord­ing to her, ac­tiv­i­ties that re­quire hu­man em­pa­thy, per­sonal con­nect and emo­tions, like griev­ance han­dling, coach­ing, coun­sel­ing which need the ‘con­nect’ and ‘trust’ quo­tient to be es­tab­lished will re­main dif­fi­cult for AI to im­pact.


Vi­nay Desh­pande, chief peo­ple of­fi­cer, Mahin­dra & Mahin­dra Fi­nan­cial Ser­vices, speaks about AI gain­ing pop­u­lar­ity in HR for its two-fold ben­e­fits of pre­scrip­tive and pre­dic­tive pat­terns. He says lot of HR ac­tiv­i­ties in­volve co­or­di­na­tion, sched­ul­ing, re­spond­ing to repet­i­tive em­ployee queries, fol­low­ing up for clo­sure etc and AI sim­u­lates hu­man in­tel­li­gence to per­form such vol­ume-based op­er­a­tional tasks with ease. “This gives an edge to HR with re­spect to speed, ac­cu­racy and vol­ume of op­er­a­tional tasks to be per­formed. The ef­fort in­vested in op­er­a­tional tasks can now be redi­rected to strate­gic de­ci­sion-mak­ing with AI based re­sults,” says he.

Sim­i­larly, he says, AI is also lever­aged for its pre­dic­tive in­sights. “It stud­ies the past trends and pre­dicts the fu­ture oc­cur­rences of events. This helps in an­tic­i­pat­ing con­tin­gen­cies and act­ing proac­tively in the VUCA world. It can be a great tool in pre­dict­ing the at­tri­tion trig­gers for an or­ga­ni­za­tion. It can study the pro­duc­tiv­ity cy­cles of an em­ployee to iden­tify a pat­tern in the highs and lows,” says he.

He, how­ever, main­tains that while AI will evolve a lot of roles in HR, it can­not re­place the cre­ativ­ity and em­pa­thy quo­tient at­tached to HR. “The softer as­pect of HR would still re­quire a per­sonal in­ter­ven­tion of Hu­man Be­ings to un­der­stand var­i­ous be­hav­ioral pat­terns of the em­ploy­ees. All roles which have a close con­nec­tion with the psy­chol­ogy of hu­man be­ings have to be re­frained from the AI frame­work. Ex­am­ple are or­ga­ni­za­tional cul­ture, em­ployee en­gage­ment,” says he.


Like in sev­eral other do­mains, HR lead­ers can de­ploy AI-based so­lu­tions to find ob­jec­tive in­sights from raw data, feels Priti Singh, vice pres­i­dent, Hu­man Re­sources, South Asia, Master­card. For in­stance, she says it can help HR pro­fes­sion­als iden­tify the most rel­e­vant can­di­dates out of a heap of as­pi­rants’ re­sumes. “This is not same as us­ing spe­cific pre-de­fined met­rics to short­list the re­sumes. It means mak­ing de­ci­sions that re­flect hu­man in­tel­li­gence. While re­sume screen­ing is just an ex­am­ple, such in­sights can be im­ple­mented in ev­ery HR func­tion such as ad­min­is­tra­tion, tal­ent man­age­ment, per­for­mance re­views, learn­ing & devel­op­ment and so on.”

She also con­curs that AI can only as­sist a hu­man HR pro­fes­sional in mak­ing the fi­nal de­ci­sion. “We will al­ways need a hu­man be­ing who brings ex­pe­ri­ence and cog­ni­tive abil­i­ties to make strate­gic de­ci­sions, she as­serts.

Ajith Kumar K.K., ex­ec­u­tive vice pres­i­dent & CHRO, Fed­eral Bank, main­tains that AI is the in­evitable next step for high-growth com­pa­nies, and no one dis­putes its abil­ity to drive ef­fi­cien­cies, scale and ef­fec­tive­ness. He says the topic of ‘AI in Hu­man Re­sources’ evokes ex­cite­ment and the ex­cite­ment is be­cause of the im­pact it can bring to the in­no­va­tion land­scape in HR. The cov­er­age of im­pact is huge, rang­ing from rein­vent­ing em­ployee ex­pe­ri­ence to data driven de­ci­sion mak­ing.

“For us in Fed­eral Bank, strength of the busi­ness case, cer­tainty of the busi­ness need, the so­lid­ity of value prepo­si­tion and surety of the ben­e­fits lay the foun­da­tion of em­brac­ing tech­nol­ogy, be it AI, RPA or ML. HR should be pru­dent in mod­el­ling the busi­ness need and break down the busi­ness chal­lenge to strate­gic ob­jec­tives to pick the right tech­nol­ogy. I must say we have an in­ter­est­ing case study of AI in re­cruit­ment, chris­tened as

Fed Re­cruit. With the suc­cess of Fed Re­cruit, we feel AI is promis­ing in HR,” he says.

He points out that the re­cruit­ment strat­egy at the bank fol­lows the brand’s core theme of ‘dig­i­tal at the fore; hu­man at the core’ by cre­at­ing a ro­bust re­cruit­ment frame­work that aligns cul­ture, tal­ent, or­ga­ni­za­tional struc­ture and pro­cesses through tech­nol­ogy. “Our fo­cus is be­yond just pro­vid­ing the hu­man touch to the can­di­dates, but more on cre­at­ing a can­di­date ex­pe­ri­ence sim­i­lar to a de­light­ful cus­tomer ex­pe­ri­ence through new tech­nolo­gies. Fed Re­cruit is de­signed to help tal­ent ac­qui­si­tion teams reimag­ine peo­ple and tal­ent pro­cesses so as to at­tract the best tal­ent, build stronger teams, and en­hance the can­di­date ex­pe­ri­ence. With the help of this AI pow­ered data-driven plat­form and HR an­a­lyt­ics, ev­ery step of hir­ing can now be mea­sured and tracked. The data pre­dic­tive hir­ing tech­nol­ogy lays the foun­da­tion of a strate­gic plan that helps to make re­li­able de­ci­sions while hir­ing tal­ent,” says he. And AI can cer­tainly do some data pro­cess­ing and de­ci­sion mak­ing faster.


Vish­wesh Pad­man­ab­han, part­ner and head, Dig­i­tal Con­sult­ing, KPMG In­dia, un­der­lines the fact that any trans­ac­tional ac­tiv­ity across the en­tire em­ployee life­cy­cle that HR as a func­tion un­der­takes - be it re­lated to re­cruit­ment, on­board­ing, tal­ent man­age­ment or sep­a­ra­tion - is be­ing grad­u­ally au­to­mated and driven through bots. “Whilst it’s only the tip of the en­tire scope of work that could be au­to­mated within HR, in the cur­rent busi­ness sce­nario, AI-driven trans­ac­tional ser­vices seems to be gain­ing the max­i­mum trac­tion,” he opines, point­ing out that this in­cludes chat­bots as a tool to drive pre­lim­i­nary HR query res­o­lu­tions mech­a­nism for em­ploy­ees.

“Given the suc­cess of bots in han­dling ba­sic trans­ac­tional ac­tiv­i­ties, more ma­ture firms have now also started to ex­per­i­ment with the idea of lever­ag­ing bots/AI to drive end-to-end HR pro­cesses. Cases in point are the re­cruit­ment AI tools that keep scan­ning a firm’s tal­ent ac­qui­si­tion sys­tem for new job post­ings ba­sis the job re­quire­ments, crawl through the re­cruit­ment sites to iden­tify the most suit­able can­di­da­tures and then to short­list and in­ter­view the ones short­listed. This helps shorten screen­ing time per can­di­date by ap­prox. 60-70% and in other re­cruit­ment ar­eas as well,” he says.

He men­tions that some or­ga­ni­za­tions are also ex­plor­ing/im­ple­ment­ing ‘hu­manoids’ to help achieve these ob­jec­tives. HR hu­manoids struc­turally rep­re­sent a hu­man and with AI they are able to for­mu­late the ba­sic con­nect and help ad­dress the sim­ple queries (eg, em­ploy­ees record de­tails, on­board­ing de­tails, train­ing de­tails, is­sue rais­ing and track­ing etc) of the work­force with speed and agility.


Po­duri of DBS Bank In­dia says the bank in the last year it­self be­yond dou­bling its head­count, had in­creased its geo­graph­i­cal spread and di­ver­sity count (in gen­der, cul­ture, and multi­gen­er­a­tion). In such cir­cum­stances, he says there are sev­eral in­stances where speed is im­por­tant for the HR team.

“Pay­roll pro­cess­ing is some­thing where speed and ac­cu­racy are im­por­tant. Noth­ing can set an em­ployee off more than a de­layed pay­check or less salary cred­ited to the ac­count. The whole can­di­date to new joiner ex­pe­ri­ence is speed-de­pen­dent, ow­ing to the bulk of data pro­cess­ing and the num­ber of peo­ple in need of at­ten­tion at the time. There is noth­ing worse in terms of ex­pe­ri­ence for a new joiner, if they have to go through a labyrinth of te­dious pa­per­work. In the in­ter­est of our tech­nol­ogy cen­ter, hackathons are very ef­fec­tive in meet­ing our hir­ing tar­gets. These events draw huge crowds, and the com­pany can cap­i­tal­ize on the qual­ity and quan­tity of hires at the same time. Speed again is cru­cial, and au­to­ma­tion is the so­lu­tion,” he says.

Desh­pande of Mahin­dra & Mahin­dra Fi­nance em­pha­sizes that when us­ing AI to the com­pany’s ad­van­tage, there is need to feed the ap­pro­pri­ate al­go­rithm to de­rive the de­sired out­put. “Or­ga­ni­za­tions to­day are iden­ti­fy­ing pro­cesses which are a low value ad­di­tion but high on trans­ac­tion and cre­at­ing al­go­rithms to au­to­mate these pro­cesses. This will boost the speed of ex­e­cu­tion with ac­cu­racy, and also re­flect un­der­ly­ing faults in the process, if any. These al­go­rithms, or ‘bots’ in other words, how­ever, need to be trained by hu­mans. For in­stance, the in­tel­li­gence part is lim­ited to AI’s abil­ity to match the JD and the qual­i­fi­ca­tions; rest of the in­ter­view and se­lec­tion process is hu­man. So, it is a ‘Ro­botic Process Au­to­ma­tion’,” says he.


What is the pref­er­ence for HR pro­fes­sion­als - cloud-based so­lu­tions or premise based so­lu­tions?

Vish­wesh Pad­man­ab­han says cloud is prov­ing to be a plat­form for in­no­va­tion. “Many new and in­no­va­tive cloud-based so­lu­tions are be­ing of­fered to­day and quite a few so­lu­tions are from new star­tups. A large part of the in­no­va­tion will be seen through in­no­va­tive tech­nol­ogy plat­forms on the cloud and there­fore, cloud-based so­lu­tions will see far more mo­men­tum,” says he.

Desh­pande avers both of­fer ap­peal­ing AI So­lu­tions. And there are star­tups which have mas­tered the ca­pa­bil­ity of of­fer­ing more ap­peal­ing AI so­lu­tions for spe­cific prob­lems, and a higher de­gree of cus­tomiza­tion of the so­lu­tions.

He adds: “A key dif­fer­ence be­tween the 2 of­fer­ings is of fi­nan­cial. Cloud-based so­lu­tions are treated as op­er­a­tional ex­pense (OpEx) be­cause they are rented monthly. In con­trast, premise-based so­lu­tions are treated as cap­i­tal ex­pense (CapEx), be­cause

they are pur­chased just once. There­fore, cloud-based so­lu­tions are pre­ferred; since you pay for the ser­vice.”

Po­duri paints a sce­nario where ev­ery­one is putting in a lot of in­no­va­tion when it comes to AI and there is cut-throat com­pe­ti­tion. And ev­ery­one is of­fer­ing cus­tomiza­tion.

“There are the star­tups (Domino Data lab, Quid, Data Ro­bot to name a few); then there are the big play­ers (Google, Ama­zon, Mi­crosoft, the usual). The choice comes down to busi­ness re­la­tions un­less some­one in the mar­ket is of­fer­ing some­thing re­ally out of the box. Cloud is the sta­tus quo at the mo­ment and DBS Bank has heav­ily in­vested in it. All HR pro­cesses and data have been moved to the cloud,” he re­veals.

Priti Singh is of the view that on-premise vs cloud is again a choice based on sev­eral fac­tors, which could be spe­cific to the busi­ness or the in­dus­try.


When it comes to the as­pects of train­ing, AI has a de­fin­i­tive role. Po­duri be­lieves Learn­ing & Devel­op­ment is the space where AI can have the most pro­found im­pact. “We are look­ing at very ex­cit­ing times, where learn­ing ex­pe­ri­ences are get­ting en­riched by aug­mented re­al­ity and vir­tual men­tors. In the fu­ture, learn­ing would be­come even more fun, com­pact, pur­pose­ful, rel­e­vant, all at the same time and all made pos­si­ble by AI.

“At DBS Bank, we are al­ready de­liv­er­ing per­son­al­ized learn­ing jour­neys to our em­ploy­ees, and I feel we have just man­aged to scratch the sur­face here. As AI al­go­rithms be­come more com­plex, we are look­ing at more ex­haus­tive con­tent cu­ra­tion, re­fine­ment in the tar­geted skill set of in­di­vid­u­als, more in­ci­sive an­a­lyt­ics, and more ag­ile prac­tices,” says he.

Vishalli Don­grie feels AI is pri­mar­ily used in the area of on-job learn­ing. For ex­am­ple, spe­cial­ist work­ers, es­pe­cially those work­ing in haz­ardous en­vi­ron­ments (like crane op­er­a­tors) can be trained us­ing a vir­tual setup to re­duce health and safety risks. The other area, ac­cord­ing to her, is to in­crease pro­duc­tiv­ity and ef­fi­ciency of work­ers on the shop floor, sales force, plant op­er­a­tions etc, where con­ven­tional class­room-based train­ing mod­els are not ef­fi­cient at scale.

“Let’s take the ex­am­ple of a lead­ing stream­ing ser­vice plat­form which helps stream ‘ on-de­mand’ con­tent to the ‘pref­er­ence’ of the user. In learn­ing too, HR/L&D teams are now ex­per­i­ment­ing with ‘real time’, ‘ on-de­mand’ learn­ing, which of­fers cus­tom­ized learn­ing to the work­force matched to ca­reer pref­er­ences, thereby help­ing i mprove the over­all ca­pa­bil­ity of the or­ga­ni­za­tion on a day to day ba­sis,” she ex­plains fur­ther.”

Desh­pande points out that AI pro­duces more re­li­able and in­sight­ful re­ports in com­par­i­son to the con­ven­tional MIS re­ports. “This brings more pre­ci­sion to the Train­ing Need Anal­y­sis and the de­sired out­come of train­ing ef­fec­tive­ness. AI is able to per­son­al­ize the train­ing pro­grams based on the learn­ing styles of the user. It can pre­dict and show con­tent as per the learner’s con­sump­tion ca­pac­ity. AI also helps in cu­rat­ing con­tent from var­i­ous sources and brings it on an ag­gre­gate plat­form thereby en­rich­ing the learner’s ex­pe­ri­ence,” he says.

Preeti Singh too says AI is play­ing a cru­cial role in en­abling per­son­al­ized learn­ing ex­pe­ri­ences for ev­ery in­di­vid­ual. This ex­pe­ri­ence can be de­liv­ered through video mod­ules, blogs, vir­tual cha­t­rooms and many other dig­i­tal-first meth­ods.

The learn­ers can set the pace of their own learn­ing and en­sure an ef­fec­tive out­come, ac­cord­ing to her.


What are the pit­falls that HR should be aware of when us­ing AI?

“From how I un­der­stand,” re­sponds Po­duri, “tech­nol­ogy is not only cre­at­ing jobs, but em­ploy­ment is be­com­ing a more ful­fill­ing ex­pe­ri­ence.”

He then talks about a flip side: “What are the func­tions that are most con­ve­niently de­ployed to the ma­chine? The most mun­dane, ev­ery­day op­er­a­tions which forms the base of our vo­ca­tion. For HR pro­fes­sion­als at the on­set of their ca­reers, mak­ing re­cruit­ment calls, pro­cess­ing pa­per­work, or scan­ning through a pile of re­sumes to find the per­fect fit for a job de­scrip­tion, these are the fun­da­men­tal func­tions which form the foun­da­tion of their ca­reer. Now with AI tak­ing over, we are com­pletely phas­ing out these ba­sic func­tions and we run the risk of hav­ing a gen­er­a­tion of un­aware pro­fes­sions. Then again, with a plethora of pos­si­bil­i­ties, there is al­ways the chance of re­dun­dancy creep­ing in. AI-gen­er­ated email re­sponses are the most rel­e­vant ex­am­ple. Very of­ten these ma­chines-gen­er­ated e-mails are out of con­text, un­nec­es­sary, or mis­in­for­ma­tive. There has to be a line, be­yond which we must trust in hu­man dis­cre­tion.”

Desh­pande says we should be cau­tious of overde­pen­dence, be­cause “here the prin­ci­ple of ‘ Garbage in, Garbage out’ is also ap­pli­ca­ble. If you have not trained the sys­tem cor­rectly, it will not give the de­sired prod­uct. We should re­mem­ber that AI is still at a nascent stage and it re­quires hu­man mon­i­tor­ing.”

Vishalli Don­grie be­lieves mun­dane things like chat bots pose no big threat as they are mostly based on ex­ist­ing ‘avail­able’ in­for­ma­tion. “How­ever, in the case of en­hanced AI, for eg, when we move to pre­dic­tive anal­y­sis-based user in­ter­ac­tions or in­for­ma­tion ex­trac­tion, the power of AI is best lever­aged when there is suf­fi­ciency of in­for­ma­tion that the sys­tem, al­go­rithm can process,” she says.

She adds that in most or­ga­ni­za­tions struc­tured HR-re­lated data cap­tur­ing are

just be­ing in­tro­duced over the last few years. Thus, lack of ac­cu­rate, enough and au­then­tic data could lead to wrong de­ci­sions by AI and HR pro­fes­sion­als needs to be mind­ful of this when cre­at­ing AI-based sys­tems.

“On top of this hu­man psy­cho­log­i­cal bi­ases are an­other big pit­fall we need to be cau­tious of while de­sign­ing AI/ ML sys­tems. For ex­am­ple, a re­cruit­ment sys­tem by a large on­line ma­jor de-se­lected fe­male can­di­dates based on the data that most ex­ist­ing tech work­ers in the firm were males. Thus, it’s all about how HR frames the prob­lem, how choices are made and avail­able data/con­texts or the lack of it will sim­ply in­flu­ence what the ma­chine needs to dis­cover. This is a ma­jor pit­fall,” says she.


Vishalli Don­grie also cau­tions that em­pa­thy is a hid­den nu­clear weapon that can touch mil­lions in one go and have a far deeper im­pact than any AI sys­tem. This skill, she says, is al­ready in short­age, and HR lead­ers now need to en­sure that em­pa­thy and emo­tional in­tel­li­gence are not be­ing af­fected by AI sys­tems and ap­pli­ca­tions. She quotes re­search by Har­vard Univer­sity, which em­pha­sized that these as­pects are key pre­dic­tors of suc­cess across man­age­rial roles.

“Last but not the least, HR needs to ask the tough ques­tion and not go with the herd men­tal­ity - ‘whether ma­chines should be al­lowed to make de­ci­sions about the fu­ture of hu­man be­ings,” she adds.

Ac­cord­ing to Priti Singh, HR roles in­volve in­terac­tion with hu­man be­ings and AI can­not al­ways be the per­fect so­lu­tion. Since AI is en­tirely ob­jec­tive in its de­ci­sion mak­ing, the data in­volved in the de­ci­sion-mak­ing needs hu­man in­ter­fer­ence to im­prove de­ci­sion­mak­ing ca­pa­bil­i­ties. There­fore, AI needs to work in sync with hu­man in­tel­li­gence.

Ajith Kumar re­calls the bank’s cor­po­rate credo ‘hu­man at the core, dig­i­tal at the fore’ to em­pha­size how the com­bi­na­tion only works ef­fec­tively. “We are deeply aware of AI’s po­ten­tial as well as pit­falls. So, when we set out to tap into the in­cred­i­ble power of AI and cre­ate an HR plat­form with AI at its heart, we were driven by this credo, he adds.


HR pro­fes­sion­als gen­er­ally be­lieve AI will help trans­form HR from an ad­min­is­tra­tive role to a mis­sion crit­i­cal role. Po­duri is em­phatic stat­ing with AI tak­ing over the rou­tine ad­min­is­tra­tive stuff, it looks al­most in­evitable that HR func­tions make the shift to a mis­sion-crit­i­cal role. Job de­scrip­tions, he ex­pects, will in­creas­ingly fo­cus on spe­cial­iza­tion within func­tions. “We al­ready have in­di­vid­u­als in our or­ga­ni­za­tion who are fo­cused on busi­ness ad­vi­sory, jour­ney think­ing, em­ployee ex­pe­ri­ence and en­gage­ment, em­ployer brand­ing. Now, these are func­tions which are defin­ing the or­ga­ni­za­tion of the fu­ture, and as we re­al­ize the fu­ture these func­tions be­come mis­sion-crit­i­cal.”

Vish­wesh Pad­man­ab­han says HR has al­ways been on the jour­ney of transition­ing away from an ad­min­is­tra­tive role to a more strate­gic role, and with the ad­vent of AI, this jour­ney will only be ex­pe­dited. “The ‘new nor­mal’, says he, “en­com­passes con­stant dis­rup­tion, game-chang­ing trends and per­pet­ual trans­for­ma­tion in the work­force and ways of work­ing (eg, gig econ­omy, mil­len­nial work­force, re­mote work­ing etc.) and in such an age, the role of HR has be­come more piv­otal and crit­i­cal than ever”

Ajith Kumar be­lieves fo­cus should be on an en­tity’s HR strat­egy. “We be­lieve that the best tech­nol­ogy plat­form for an HR func­tion will be the one that best aligns with the HR strat­egy that sup­ports busi­ness goals. Hasty de­ci­sions in the bells and whis­tles of tech­nol­ogy might not help. Needs are unique to or­ga­ni­za­tions and due dili­gence should be ex­er­cised be­fore pick­ing AI or any other bleed­ing edge tech­nol­ogy. In re­cruit­ment we, ex­er­cised strict vigil be­fore adopt­ing AI. Once the mind­set was cre­ated af­ter mod­el­ling the need, it be­came easy for us to reach the so­lu­tion,” says he.

Vi­nay Desh­pande fore­sees all rou­tine tasks would be taken care of by AI and bot work­ers and HR will be more fo­cused on de­ci­sion-mak­ing and course cor­rec­tion. “HR will, ad­di­tion­ally, have to cater to train­ing the bots and en­sur­ing their align­ment to the HR pro­cesses. This would re­quire the HR to up­skill them­selves to train the bots and en­sure the us­age of rel­e­vant al­go­rithms. AI can­not take de­ci­sions for you; it can only help you make those de­ci­sions. AI will in­duce a par­a­digm shift from an ad­min­is­tra­tive role to a con­sul­ta­tive role for HR for sure,” says he.


Po­duri be­lieves there are sev­eral av­enues within HR func­tions where AI has been im­ple­mented to mon­i­tor t he rec­om­men­da­tions and de­ci­sions emerg­ing from AI. “Di­ver­sity in func­tions means that the mod­er­a­tion process needs to pro­tect the essence of that func­tion. Fre­quent sim­u­la­tion runs and pro­vi­sions for a hu­man over­ride are some mod­er­a­tion tech­niques that I can think of. To elim­i­nate er­rors in pro­ce­dure, ul­ti­mate de­ci­sion-mak­ing needs to be vested on hu­mans, sup­ported by an ex­haus­tive feed­back mech­a­nism,’ says he.

Desh­pande feels the rec­om­men­da­tions and de­ci­sions have to be con­stantly re­viewed. There has to be a pilot pro­ject, there has to be a trial of the pilot pro­ject, and there has to be an anal­y­sis on the ba­sis of how the work is go­ing on, he says. Be­yond this, there has to be a hu­man in­ter­ven­tion to val­i­date the rec­om­men­da­tions made be­fore it is scaled up.

“Hu­mans also have to ex­plore any al­ter­nate best method of per­form­ing the same task which the AI might not be pro­grammed to con­sider. There will also be a cost ben­e­fit anal­y­sis at­tached to the

fi­nal so­lu­tion that has been con­sid­ered and then build­ing a cul­ture to em­brace AI, es­pe­cially in In­dia. But that will be­come eas­ier when more bot work­ers come into use,” he elab­o­rates.


Vishalli Don­grie ar­gues that there needs to be a re­view mech­a­nism in the form of a steer­ing com­mit­tee with rep­re­sen­ta­tion from HR as well as other func­tions, to suc­cess­fully drive out­comes and re­view such rec­om­men­da­tions from AI projects. Such an ar­range­ment will al­low the lead­er­ship to work to­gether dur­ing all phases – ideation, re­al­iza­tion and change adop­tion, she adds.

DBS Bank has al­ready cov­ered some ground in redefin­ing job de­scrip­tions in the light of us­ing AI for HR func­tions and the next step would be to push for more com­pre­hen­sive spe­cial­iza­tion. Po­duri is con­fi­dent this would give rise to re­fined skillsets and the in­clu­sion of a plethora of new av­enues into the main­stream. Cul­tural ar­chi­tect, emo­tional an­a­lyst, pol­icy strate­gist, change man­ager and Org Dev man­ager are some of the roles, where the func­tions are al­ready un­der im­ple­men­ta­tion in the bank, he re­veals.

Desh­pande ex­pects the job de­scrip­tions will evolve from the cur­rent con­text. There will be need for more techno HR pro­fes­sion­als, who will ap­pre­ci­ate tech­nol­ogy, adopt the same in the HR sce­nario and train the al­go­rithms (bots) to en­hance the HR pro­cesses. Some­one pro­fi­cient in data science, who can eval­u­ate the qual­ity of rec­om­men­da­tions made by the AI tool, will be the pref­er­ence, ac­cord­ing to him.


For Vish­wesh Pad­man­ab­han how­ever, the head­line re­quire­ment still re­mains the same and it all comes back to man­ag­ing peo­ple. “What emerg­ing tech­nol­ogy trends will do is give HR bet­ter tools and means to man­age and ‘de­light’ its peo­ple. So, in essence, the role and suc­cess for HR lead­er­ship will still be de­fined through same met­ric. How­ever, mid­dle and ju­nior man­age­ment roles may be­come less trans­ac­tional and more strate­gic and data driven – with need to en­able bet­ter em­ployee ex­pe­ri­ence, en­gage­ment, en­able­ment and ed­u­ca­tion,” he elab­o­rates.

Preeti Singh be­lieves that AI is im­pact­ing the job roles and ex­pec­ta­tions of the or­ga­ni­za­tion from the em­ploy­ees. Since most of the func­tions are now evolv­ing and in­cor­po­rat­ing new tech­nolo­gies, em­ploy­ees need to have the abil­ity to work with these changes, she says. “Go­ing for­ward, they will be less re­quired for it­er­a­tive func­tions which a ma­chine can learn and per­form faster. The real jobs for HR pro­fes­sion­als will be to add hu­man in­tel­li­gence on top of the in­sights gen­er­ated by AI-based plat­forms and make the right de­ci­sions, says she.


One model of­ten pro­posed for us­ing AI in an or­ga­ni­za­tion is that it serves as an in­tel­li­gent as­sis­tant to in­di­vid­u­als. What are the pros and cons of this model?

Po­duri be­lieves this is an in­ter­est­ing propo­si­tion. Says he: “With vir­tual home as­sis­tance tak­ing the mar­ket by storm, we are al­ready some way out there. How­ever, the propo­si­tion of a per­sonal in­tel­li­gence sys­tem does come with its pros and cons. The idea pro­poses to raise the per­for­mance stan­dards of the whole or­ga­ni­za­tion, as in­stant, cus­tom­ized as­sis­tance would al­low the in­di­vid­u­als to fo­cus on as­pects of their ca­reer which are as­pi­ra­tional and more ful­fill­ing to them. At the same time, there is the stark pos­si­bil­ity of over depen­dence on ma­chine, and a cut away from hu­man con­nec­tions. We need to be very thought­ful and have a bal­anced ap­proach on what and how we im­ple­ment AI,” he says.

Ajith Kumar be­lieves the ben­e­fits are nu­mer­ous. It can speed up ac­tiv­i­ties, it can help make faster de­ci­sions, many small as­pects of the daily work­ing life can be han­dled with ease and in a more or­ga­nized and co­or­di­nated mat­ter.

He points out that the Fed Re­cruit mo­bile ap­pli­ca­tion is in­te­grated with NLPbased chat­bot that not only turns the top­down em­ployer-ap­pli­cant mono­logue into a hor­i­zon­tal di­a­logue, but en­hances the en­tire can­di­date ex­pe­ri­ence.

“By em­ploy­ing NLP based chat­bot so­lu­tion to com­ple­ment our re­cruit­ment team, we elim­i­nated the repet­i­tive tasks of an­swer­ing phone calls. The well-op­ti­mized chat­bot com­mu­ni­cates only the es­sen­tials and does not over­whelm the can­di­date. The chat­bot rep­re­sents the or­ga­ni­za­tion when it is com­mu­ni­cat­ing with the can­di­date, so, from a brand­ing point of view, it is a per­fect way of em­ployer brand build­ing,” says he.

Desh­pande re­it­er­ates that while AI will ease the process of HR ex­e­cu­tion, it will not be able to re­place the Em­pa­thy, Cre­ativ­ity and Strate­gic Think­ing in an or­ga­ni­za­tion. Ac­cord­ing to him the pros of this propo­si­tion are:

Speed: AI pro­duces di­ag­nos­tics on real time data faster than man­ual con­ven­tional MIS re­ports. It will au­to­mate all the non-value ac­tiv­i­ties like co­or­di­na­tion, sched­ul­ing, data min­ing, data col­la­tion etc.

Cost: AI will help stream­line lot of pro­cesses, thereby keep­ing the HR team Lean and re­duc­ing the man­power cost.

Qual­ity: Since the an­a­lyt­ics is pro­grammed, there is no scope for de­vi­a­tion, thus con­trol­ling any er­rors.

Quan­tity: AI can eas­ily han­dle large vol­umes of data from var­i­ous sources and de­rive re­sults from it, which would be re­stricted in case of hu­man driven an­a­lyt­ics.

And the cons are:

Re-skilling and up-skilling the work­force: While AI will not cut down jobs, it will re­quire em­ploy­ees to up-skill or re-skill them­selves in or­der to adapt to the AI-based mod­els. Some­one who fails to

match pace with the evolv­ing na­ture of HR will have to part ways even­tu­ally.

Soft skills: AI will not be able to repli­cate the same de­gree of em­pa­thy to­wards its em­ploy­ees as an HR per­son. For ex­am­ple, if an in­di­vid­ual has to nudge you for a re­minder, he/she will do it in a hu­man way, keep­ing your men­tal and phys­i­cal health in mind. The bot will do it in a very me­chan­i­cal way. It won’t con­sider such par­al­lel fac­tors.

Vishalli Don­grie be­lieves AI can serve as in­tel­li­gent as­sis­tant to in­di­vid­u­als in the ar­eas like sales, mar­ket­ing, HR & re­cruit­ment and cus­tomer ser­vice. She says some of the most mun­dane of­fice tasks like writ­ing emails, sched­ul­ing meet­ings, tak­ing notes, mak­ing travel ar­range­ments etc can be han­dled by such as­sis­tants thus help­ing the em­ploy­ees save on time. This will work quite well with the more tech savvy younger gen­er­a­tion.

“The cons,” ac­cord­ing to her, “could be ‘ad­dic­tion’ - be­com­ing slaves to tech­nol­ogy and also los­ing out on the hu­mane el­e­ments of the task - a com­plete trans­for­ma­tion of how we think, live and feel. AI as in­tel­li­gent as­sis­tant is mostly re­ac­tive in na­ture and not proac­tive. Loss of cre­ativ­ity of the hu­man mind could be a ma­jor draw­back for this model. The worst part is AI learns from us and hence if we em­brace the wrong be­hav­iors, then this is what AI will start be­liev­ing as nor­mal and re­spond in that man­ner.”


Vish­wesh Pad­man­ab­han em­pha­sizes that in­creas­ing use of AI in HR will cre­ate newer po­si­tions and roles in the HR depart­ment. “With AI play­ing a big­ger role in driv­ing HR pro­cesses, HR will have more peo­ple data to an­a­lyze than ever be­fore - which will re­quire HR to be­come more an­a­lyt­i­cal and build ca­pa­bil­ity around cre­ative de­sign think­ing to re­shape the way em­ploy­ees are ser­viced in fu­ture. In ad­di­tion, we could also see tech roles or com­pe­ten­cies like data min­ing be­ing added to the HR func­tions. There could be roles like chief eth­i­cal of­fi­cer, a role that en­sures tech­nol­ogy is used in an eth­i­cal and hu­mane way, and hu­man re­sources tech trainer roles, etc,” he says.

Po­duri says with the in­creas­ing use of AI, peo­ple can de­velop their niche and spe­cial­ized skills and with more fo­cus on spe­cial­iza­tion, it is only a mat­ter of time be­fore the con­tri­bu­tions made by in­di­vid­u­als re­quire the cre­ation of new po­si­tions.

Ajith Kumar says Fed­eral Bank has po­si­tioned AI in re­cruit­ment as an en­abler, not just safe­guard­ing ra­pid­ity, but rein­vent­ing the hu­man in­ter­ven­tion. “For us, the HR in­ter­ven­tion is now re­quired in fewer ways for the op­er­a­tional ar­eas of tal­ent ac­qui­si­tion and is far more im­per­a­tive in the an­a­lyt­i­cal and in­tel­lec­tual as­pects. The data pre­dic­tive hir­ing tech­nol­ogy lays the foun­da­tion of a strate­gic plan that helps to make re­li­able de­ci­sions while hir­ing tal­ent. The en­gage­ment fo­cused tal­ent ac­qui­si­tion can scale up to very large tal­ent pools while still pre­serv­ing the qual­ity of can­di­date ex­pe­ri­ence,” he ex­plains.

Desh­pande also agrees that with the evolv­ing role of AI in HR, there will be techno HR pro­fes­sion­als. “You will need a spe­cial skill-set to train the AI, and to use the re­lated ap­pli­ca­tions in a man­ner that is pro­duc­tive. There will also be a need to do a Cost-Ben­e­fit Anal­y­sis (CBA), and there­fore, spe­cial­ists who can do that job ef­fi­ciently, will be wel­comed,” he says.

Fi­nally, what is their take on the mostoften stated fear that AI will take away jobs?

Priti Singh states it is a fact that the fear that AI and ro­bots will au­to­mate many jobs has been grow­ing. She points out to stud­ies car­ried out in the US, which have pre­dicted that 40%-50% of jobs in the coun­try could be au­to­mated in the next 20 years, and many peo­ple are wor­ried that they could be forced into re­tire­ment. “How­ever, the way hu­mans work has al­ways been chang­ing. Ever since the First In­dus­trial Rev­o­lu­tion to this day, when we talk about in­dus­try 4.0, peo­ple have not lost jobs be­cause of tech­nol­ogy. They have in­stead got a higher level of func­tions and had to up­skill ac­cord­ingly,” she avers.

Po­duri says em­phat­i­cally: “That is not in the very least true. I be­lieve that tech­nol­ogy is cre­at­ing more ful­fill­ing jobs for the peo­ple, where they can dive deeper into their sub­jects of in­ter­est. How­ever, they must be open to change, be stoked at the op­por­tu­nity to learn, un­learn and do it again.”

And Vishalli Don­grie points out that we have seen this ques­tion com­ing up with dis­rup­tive tech­nolo­gies in the past as well, and this is not the first time that a new tech­nol­ogy has im­pacted jobs. “This hap­pened when elec­tric­ity came in and re­placed steam power, com­put­er­i­za­tion came in to re­place man­ual pro­cesses and many oth­ers. What we have learnt from sim­i­lar tech­nol­ogy changes in the past is that the ini­tial pe­riod re­quires ad­just­ment, but af­ter that time pe­riod, tech­nol­ogy cre­ates new jobs and roles for em­ploy­ees, which are of­ten hard to en­vi­sion at the start of the jour­ney. How­ever, when it comes to AI in­te­gra­tion to­day, the key­word is very much ‘aug­men­ta­tion’ - the idea that AI ma­chines will help us do our jobs more ef­fi­ciently, rather than re­place us. A key idea is that they will take over the mun­dane as­pects of our role, leav­ing us free to do what hu­mans do best - tasks which re­quire cre­ativ­ity and hu­man-to-hu­man in­terac­tion.”

Desh­pande, how­ever, ad­mits that some jobs will def­i­nitely be af­fected, es­pe­cially the rou­tine jobs, trans­ac­tional jobs and jobs which are low value-based. “For ex­am­ple, nowa­days, hardly any­one uses a phys­i­cal dictionary to find the mean­ing of a word; in­stead they Google it. How­ever, cre­ativ­ity is the one as­pect with re­gards to AI that is pos­si­ble, but is still re­stricted,” he high­lights.

Kishore Po­duri un­der­lines the fact that speed is cru­cial in HR on ac­count of sev­eral par­al­lel ac­tiv­i­ties and au­to­ma­tion - ba­si­cally use of AI - is the so­lu­tion

Vishalli Don­grie em­pha­sizes that ac­tiv­i­ties that re­quire hu­man em­pa­thy, per­sonal con­nect and emo­tions will re­main dif­fi­cult for AI to im­pact

Vi­nay Desh­pande avers AI is gain­ing pop­u­lar­ity in HR for its two-fold ben­e­fits of pre­scrip­tive and pre­dic­tive pat­terns

Vish­wesh Pad­man­ab­han speaks about how in­creas­ing use of AI in HR will cre­ate newer po­si­tions and roles in the HR depart­ment

Priti Singh in­sists we will al­ways need a hu­man be­ing who brings ex­pe­ri­ence and cog­ni­tive abil­i­ties to make strate­gic de­ci­sions

Ajith Kumar be­lieves the cov­er­age of im­pact of AI in HR is huge, rang­ing from rein­vent­ing em­ployee ex­pe­ri­ence to data driven de­ci­sion mak­ing

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.