We use cookies on this website. By continuing to use this site without changing your cookie settings, you agree that you are happy to accept our privacy policy and for us to access our cookies on your device.
Transcript/ScriptLOGON AI MASKING ACCENTS
HEADLINE: LogOn: Accent Masking Software Aims to Smooth Call Center Interactions
TEASER: AI tool replaces nonnative English speakers’ accents with more 'neutral’ voice
PUBLISHED AT: 9/26/23 at 9:50a
BYLINE: Matt Dibble
CONTRIBUTOR:
DATELINE: Palo Alto, CA
VIDEOGRAPHER: Matt Dibble
VIDEO EDITOR: Matt Dibble
ASSIGNING EDITOR: Stearns
SCRIPT EDITORS: Stearns, DLJ
VIDEO SOURCE (S): VOA original, AP, Reuters, Storyblocks, Sanas AI,
PLATFORMS (mark with X): WEB __ TV _x_ RADIO_ x_
TRT: 1:58
VID APPROVED BY: Reifenrath
TYPE: TVPKG
EDITOR NOTES:))
((INTRO)): A California tech startup is using artificial intelligence to mask the accents of call center workers. The goal is to reduce the ill feelings workers encounter when taking calls from people who have trouble understanding them. Matt Dibble has the story.]]
((NARRATOR))
In the global economy, a call to a customer service center will likely be answered by someone in a faraway part of the world. Call center workers can train for months to effectively communicate with customers whose native language and culture differ from theirs.
((NARRATOR))
Despite all the training, an agent’s accent can trigger biased reactions from a caller, says developer Sharath Narayana.
((Sharath Keshava Narayana, Sanas AI Co-founder))
“Now I'm connected to somebody that I can't relate to or his voice that I cannot relate to. Can this person actually solve my problem?’”
((NARRATOR))
Narayana is a co-founder of California-based Sanas AI, whose technology can adjust the way a speaker sounds with the goal of making an accent more relatable. A call center agent from the Philippines demonstrated.
((SOT))
“Hi my name is Iggy. I’m from Parañaque.”
“Hi my name is Iggy. I’m from Parañaque.”
((NARRATOR))
The agent speaks normally into a microphone and the AI system detects phonemes, which are the sound components of words.
((Sharath Keshava Narayana, Sanas AI Co-founder))
”What we're essentially doing in real time is ((Courtesy: Sanas AI)) matching those phonemes to how that phoneme should sound in an American accent.” ((end courtesy))
((NARRATOR))
Discrimination and abuse by callers are among the factors leading to high turnover among call center workers. Narayana himself once worked in such a center, and although he doesn’t excuse the behavior of callers, he says Sanas makes the job easier for agents.
((Sharath Keshava Narayana, Sanas AI Co-founder))
“If there is a technology solution to it where I don't have to change the way I sound, but the other person can understand me very, very clearly, ((Courtesy: Sanas AI))
why isn't it a beautiful thing?”
((SOT)) “I was able to listen to one of my calls using Sanas and wow, it looks like English is my first language." ((end courtesy))
((NARRATOR))
Narayana says that 97% of agents with an optional Sanas system installed are using it every day.
((Matt Dibble, VOA News, Palo Alto, California))
NewsML Media TopicsArts, Culture, Entertainment and Media
NetworkVOA
Location (dateline)
Palo Alto, California
Embargo DateSeptember 26, 2023 14:12 EDT
Byline
Matt Dibble, VOA News
Brand / Language ServiceVoice of America - English