Look Who’s Talk­ing

The tech­nol­ogy be­hind Ama­zon’s Alexa voice ser­vice will con­verse with a client, un­der­stand the re­quest and de­liver what’s needed.

Financial Planning - - TECH ZONE - BY SULE­MAN DIN


Ad­vi­sors won­der­ing how Ama­zon will en­ter wealth man­age­ment should look to its cloud com­put­ing arm, Ama­zon Web Ser­vices, which is pitch­ing the nat­u­ral-lan­guage-pro­cess­ing, ma­chine-learn­ing brain be­hind its voice in­ter­ac­tive ser­vice to wire­houses, bro­ker-deal­ers and robo ad­vi­sors.

Al­ready, UBS has part­nered with Ama­zon to en­able call­ers, both clients and non­clients, to get an­swers to fi­nan­cial and eco­nomic ques­tions through Alexa, the fe­male voice as­so­ci­ated with the Ama­zon Echo de­vices.

LPL’S CEO Dan Arnold gave a thumbs-up to the idea of in­te­grat­ing Alexa into the work­flow of ad­vi­sors. Fi­delity is test­ing Ama­zon Web Ser­vices ap­pli­ca­tions, and Cap­i­tal One is al­ready a cus­tomer. Bet­ter­ment is us­ing 20 dif­fer­ent ap­pli­ca­tions as well.

Still, plan­ners might be for­given for won­der­ing what use a cylin­dri­cal lit­tle elec­tronic de­vice could be to their prac­tice. In fact, there are a va­ri­ety of ways that the tech­nol­ogy un­der­pin­ning Alexa is likely to spread through­out wealth man­age­ment, in­dus­try ex­perts say, as it au­to­mates client in­ter­ac­tion and even ad­vice re­quests.

“Many im­por­tant plat­form shifts start out look­ing like a toy and then be­come the pri­mary way that the younger gen­er­a­tion com­mu­ni­cates,” says Lex Sokolin, direc­tor of fin­tech strat­egy at Au­ton­o­mous Re­search. “Look at mo­bile apps, so­cial net­works and now con­ver­sa­tional in­ter­faces and vir­tual re­al­ity. It’s those com­pa­nies that in­no­vate at the edges that can cap­ture new cus­tomers.”


In­deed, at the cav­ernous Jav­its Con­ven­tion Cen­ter in mi­dau­gust, a room set aside for pre­sen­ta­tions on ap­pli­ca­tions of Ama­zon Web Ser­vices for the fi­nan­cial ser­vices in­dus­try was packed all day, with lines of at­ten­dees wait­ing to get in.

Ex­plain­ing to the crowd how Alexa’s cloud-based brain, Ama­zon Lex, can be used by fi­nan­cial firms, Felix Can­de­lario, a so­lu­tions ar­chi­tect at Ama­zon Web Ser­vices, didn’t soft-pedal its ca­pa­bil­i­ties. “Nat­u­ral lan­guage pro­cess­ing is the new rocket science,” he said.

That’s the abil­ity to have a con­ver­sa­tion with a client, as any hu­man rep­re­sen­ta­tive would, and still un­der­stand any re­quest cor­rectly and de­liver the ser­vice they have asked for.

The ex­am­ples demon­strated were sim­ple but col­lo­quial: “Cool. Can I trans­fer 100 bucks from my check­ing to sav­ings ac­count?” But firms can in fact use the plat­form to han­dle more com­pli­cated re­quests, Ama­zon says, not­ing the sys­tem is con­stantly learn­ing.

So, how could Alexa an­swer this ques­tion: “I have a 401(k) and I want to roll it over into an IRA, which one is best?” Hany­bal Ja­joo, global ac­count so­lu­tions ar­chi­tect at Ama­zon Web Ser­vices, ex­plains the re­quest would go through the chain of con­ver­sa­tional chat­bot, to Ama­zon Lex and then an event-driven server­less plat­form (a com­put­ing plat­form) called AWS Lambda. It would be matched up with the cus­tomer’s in­for­ma­tion based on his or her pro­file.

“It would pro­vide the op­por­tu­nity to prompt the user for ad­di­tional in­for­ma­tion,” Ja­joo says. “The chat­bot would then pass the con­text of the user to some sort of en­gine that

can per­form the rec­om­men­da­tion in real time. With this in mind, the sce­nario would be pos­si­ble.”

There are other uses, says Brian Shen­son, for­mer vice pres­i­dent of ad­vi­sor tech­nol­ogy at Sch­wab. Alexa is al­ready do­ing them for its re­tail users now, he points out: sched­ul­ing meet­ings and in­form­ing them of up­com­ing events or needs.

“You’re driv­ing to work and you sud­denly get an idea you need to in­te­grate into your dis­cus­sion with your client,” Shen­son says. “You have the fa­cil­ity to do that with­out hav­ing to write a note. It can be much more pow­er­ful in the de­liv­ery of client en­gage­ment and com­mu­ni­ca­tions.”


Shen­son, who now runs his own con­sult­ing firm in the Bay Area, sees wealth man­age­ment firms cau­tiously ap­proach­ing the use of nat­u­ral lan­guage pro­cess­ing and ma­chine learn­ing.

“I imag­ine larger firms with sig­nif­i­cant risk aver­sion will strug­gle to em­brace it. It’s Day 1; it’s too black box. They don’t un­der­stand how it works. This is sim­i­lar to how robo ad­vi­sors played out. At first, firms didn’t un­der­stand how they worked be­hind the scenes. Now, no one thinks twice about the al­go­rithms in the robo plat­form.”

In ad­di­tion to client en­gage­ment, key wealth man­age­ment ap­pli­ca­tions for the tech­nol­ogy in­clude ad­vi­sor enablement, in­vest­ment ideation and risk man­age­ment, says Will Trout, head of wealth man­age­ment re­search at Ce­lent.

“The ser­vice could be ex­tended even to de­liver ad­vice,” Trout says. “So it’s like a chat­bot, but re­la­tion­ship-cen­tered, and not trans­ac­tional, like most of the re­tail bank­ing chat­bots. Plus, it has very strong nat­u­ral lan­guage pro­cess­ing and gen­er­a­tion ca­pa­bil­i­ties. Don’t un­der­es­ti­mate Ama­zon.”

Other firms are of­fer­ing ma­chine learn­ing-en­hanced ser­vices too, such as Sales­force’s re­cently un­veiled CRM for ad­vi­sors. There are in­dus­try ef­forts to de­velop con­ver­sa­tional in­ter­faces also, such as Bank of Amer­ica’s Erica. But Ama­zon has an edge, Shen­son says, be­cause the base for im­prov­ing Alexa ex­tends into Ama­zon’s cus­tomer base, which is far beyond one in­dus­try.

“They ben­e­fit from mil­lions of in­ter­ac­tions,” he says. “And when they are widely em­braced by the con­sumer mar­ket first, it lends more cred­i­bil­ity than if it they tried to emerge first as a busi­ness tech­nol­ogy.”


Such of­fer­ings raise the ques­tion of what’s the end game for Ama­zon Web Ser­vices in the fi­nan­cial in­dus­try?

Scott Mullins, head of world­wide fi­nan­cial ser­vices busi­ness de­vel­op­ment at Ama­zon Web Ser­vices, says the goal is the same for all fin­tech providers: to help in­cum­bent wealth man­agers ease into the dig­i­tal age.

“Wealth man­agers can devote more time to in­no­vat­ing and ad­dress­ing cus­tomer chal­lenges and less time wor­ry­ing about in­fras­truc­ture and tech­nol­ogy de­ploy­ment,” he says.

Mullins notes that Bet­ter­ment has off­loaded a num­ber of its in­ter­nal func­tions to the cloud ser­vices provider. “They’ve gone all in with AWS, lever­ag­ing ser­vices like Ama­zon Red­shift, AWS Lambda, AWS Data­base Mi­gra­tion Ser­vice, Ama­zon Ki­ne­sis, Ama­zon Dy­namodb, and more,” Mullins says. “To­day, they are us­ing over 20 AWS ser­vices to de­velop, test and de­ploy fea­tures and en­hance­ments on a daily ba­sis.”

Shen­son sees the Alexa pitch as the way cloud ser­vices at­tempt to gain greater adop­tion from the fi­nan­cial ser­vices in­dus­try, which still has reser­va­tions about shar­ing highly reg­u­lated client in­for­ma­tion.

But do ad­vi­sors have to think about this ser­vice be­com­ing a bridge for Ama­zon’s re­tail branch to com­pete di­rectly for clients, as robo ad­vi­sors did? Trout says no.

“They want to serve the fi­nan­cial ser­vices in­dus­try, yes. That doesn’t mean they want to en­ter it as a com­peti­tor. When you are try­ing to rein­vent space travel, who cares about re­tail bank­ing?”

“Plat­form shifts start out look­ing like a toy and then be­come the pri­mary way that the younger gen­er­a­tion com­mu­ni­cates,” says Lex Sokolin of Au­ton­o­mous Re­search.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.