October 06, 2026 07:35 pm (IST)
Follow us:
facebook-white sharing button
twitter-white sharing button
instagram-white sharing button
youtube-white sharing button
‘Crash axe’ blow fractured Smit Machchhar’s skull: Shocking details emerge from flydubai cockpit horror | ‘Vote thief’ Gyanesh Kumar: Rahul Gandhi climbs barricade, waves Tricolour during explosive INDIA bloc protest | ‘Crash axe’ blow fractured Smit Machchhar’s skull: Shocking details emerge from flydubai cockpit horror | ‘He should not have been flying’: JD Vance on Flydubai co-pilot | Israel was ready to shoot down Flydubai jet if it kept flying towards Tel Aviv: Reports | 'If law wasn't followed, we can undo it': Supreme Court's big warning on SIR process amid Gyanesh Kumar row | From TMC Rajya Sabha MP to BJP nominee in 6 months: Koel Mallick's stunning political U-turn | ‘Stomp on your head’: Omani flydubai co-pilot’s chilling posts on violence against women surface | Asian Games 2026: India beat Pakistan to clinch gold in men's cricket | flydubai cockpit attack: Deleted posts expose Omani co-pilot’s views on women, alleged extremist past
Photo Courtesy: Pixabay

SoftBank develops AI-driven voice tech to deal with angry customers

| @indiablooms | Jun 15, 2024, at 04:33 am

Japanese tech giant SoftBank Corp has introduced an innovative voice-altering phone technology to combat the growing issue of "customer harassment."

This artificial intelligence (AI)-driven solution changes the voices of even the angriest callers to sound calm and composed, thereby reducing stress levels among call-center operators, according to The South China Morning Post (SCMP).

The Tokyo-based company aims to address the increasing stress experienced by call-center staff who face frequent angry outbursts from dissatisfied customers, the report stated.

SoftBank engineers have been working on the "emotion-cancelling" system for three years, inspired by Toshiyuki Nakatani, an employee who saw a television program highlighting the verbal abuse suffered by many staff members.

“We developed ‘emotion cancelling’ in response to the social issue of customer harassment of call-centre staff and to protect them,” Nakatani was quoted as saying by SCMP.

The technology employs a two-step process. First, it uses AI voice-processing technology to identify an angry caller and extract the key features of their comments.

The second step involves integrating the acoustic characteristics of a non-threatening voice to produce a natural and calm tone.

This approach allows operators to understand the situation without being overwhelmed by aggressive intonations.

The AI system has been trained on over 10,000 items of voice data, with 10 male and female actors delivering more than 100 common phrases in various emotional tones, including shouting, accusations, and demands for apologies.

Importantly, the technology does not alter the caller’s words but significantly tones down the intonation. Callers remain unaware of the modifications to their voice.

For instance, a high-pitched female voice is automatically lowered to be less resonant, while a booming male voice is raised in pitch to sound softer.

Despite the alterations, some elements of anger are retained to ensure that operators can appropriately respond to the caller's concerns.

A recent survey conducted by the 1.8-million-member UA Zensen union, Japan’s largest industrial union, revealed that 46.8 percent of service industry workers had experienced customer harassment in the past two years. Some were so traumatized that they required counselling.

Support Our Journalism

We cannot do without you.. your contribution supports unbiased journalism

IBNS is not driven by any ism- not wokeism, not racism, not skewed secularism, not hyper right-wing or left liberal ideals, nor by any hardline religious beliefs or hyper nationalism. We want to serve you good old objective news, as they are. We do not judge or preach. We let people decide for themselves. We only try to present factual and well-sourced news.

Support objective journalism for a small contribution.