A fictional story. The characters, organisations and events are imaginary. Saloma is real.
In 30 seconds Where: a 24-hour BPO call centre in Cyberjaya serving customers in Australia, the UK, the US and Taiwan. What Saloma did: handled a three-fold call surge on the stormiest night of the year, passed complex cases to human agents with a full summary, and gave an exhausted night shift room to breathe. Why it matters: Saloma is an enterprise AI voice agent by Astiostech that takes the repetitive calls so contact centre agents can focus on the ones that need people.
Reading time: about 5 minutes
An email with a countdown
At 2:58am, Jessie Wong, head of operations at a big Cyberjaya BPO, received an email from her biggest client, an Australian online retailer called Wombat & Co.
"Last Mega Sale, customers waited 47 minutes on hold. This year's sale is in 21 days. If waits go over two minutes again, we're moving providers. No hard feelings. Cheers."
Nothing in the world sounds scarier than an Australian saying "no hard feelings, cheers."
She forwarded it to her night shift team leader, Kavin. Kavin read it, then glanced at the envelope in his bottom drawer. It said RESIGNATION LETTER. Half his team had already quit. The other half ran on Milo and prayers.
At 3:12am, Kavin called his cousin. "Farid. That AI your boss keeps talking about. Can she speak Australian?"
"She speaks English, Kavin."
"Bro, you have never heard an Australian at 3am. That is English after a long weekend."
Overseas, spiritually
"International customers!" said YB Zainal the next morning. "Saloma's first overseas assignment. I must come."
"YB, the building is in Cyberjaya. Thirty minutes away."
"But the customers are overseas. So, spiritually, we are going overseas."
He packed a jacket ("call centres are very cold, Farid, I have heard stories"), a travel pillow and his laminated card marked FOR NATIONAL EMERGENCIES ONLY. Farid packed antacids.
Mission control, powered by teh tarik
The operations floor had hundreds of headsets and a wall of clocks: Sydney, Taipei, Dubai, London, New York. Jessie had eighteen years in BPO and the calm of a woman who once handled a complaint about a lost parcel that turned out to be a live goose.
"YB, most AI demos I've seen sound like a GPS having a bad day."
"Ms Jessie, Saloma once talked a Penang uncle out of a durian prescription. Your Australians do not frighten us."
Jessie explained the real problem. The same few questions arrive all night: Where's my order? How do I return this? Why was I charged twice? "We don't have a volume problem," she said. "We have a tiredness problem."
G'day, Saloma
The first caller on Saloma's line was Dazza from Brisbane, who sounded like he was surfing during the call.
"G'day. Me parcel's gone walkabout, mate. Barbie cover, ordered a fortnight ago."
Kavin went pale. Saloma did not.
"G'day! Sorry your barbecue cover hasn't arrived. Let's find it. What's your order number?"
She found it stuck at a depot, gave him a new delivery date and texted him a tracking link.
"Too easy. You're a legend."
Two minutes, fourteen seconds. Kavin put down his fifth Milo.
The YB takes a call
By night three, the YB wanted to serve the people directly. Even if the people were in Texas.
They gave him a headset. The caller was Earl, whose cowboy boots were one size too small.
"Assalamualaikum and a very good evening. On behalf of the rakyat of Malaysia, I extend to you our warmest greetings, and assure you that your concerns are our concerns, and together, we shall..."
A long pause from Texas. "Uh. Sir? Is this... the robot?"
Kavin quietly transferred the call to Saloma, who sorted the exchange in forty seconds.
"Now THAT sounds like a real person," said Earl. "That other fella sounded like he was running for office."
The floor collapsed. The YB removed his headset with enormous dignity. "Farid. Do not post that."
Farid had already saved the recording. For historical purposes.
The very polite standoff
The next week, Jessie added a Mandarin line for a client in Taipei. The first caller, Mr Chen, opened with a beautifully polite apology for disturbing them.
Saloma replied with an equally polite apology for his inconvenience.
Mr Chen thanked her for her patience. Saloma thanked him for his. He apologised for taking her time. She assured him it was her pleasure.
Agents gathered round the monitor like it was a penalty shootout. Kavin put RM10 on Mr Chen.
Then Saloma gently moved forward: "Please allow me to arrange your exchange right away, at no cost."
Mr Chen agreed, and left a five-star rating calling her "more courteous than a five-star hotel concierge." Kavin lost his RM10 and called it the best RM10 he had ever lost.
Mega Sale night, and the storm
Day twenty-one arrived with thunder. By 11pm, flash floods had cut several roads into Cyberjaya. Half of Kavin's shift was stuck, sending photos of their cars in what looked like a small river.
The sale went live at midnight. Calls tripled within the hour. On a dashboard shared with Wombat & Co, the average wait time sat under a minute.
Saloma took the flood of "where's my order" calls. The agents who made it in took the hard ones.
At 2:40am, Mrs Patterson called from London: charged three times, card blocked by her bank, flying to Spain in six hours. Saloma recognised it needed a human and transferred her with a summary. Before Kavin said hello, he could see everything.
"Mrs Patterson, I can see you've been charged three times and you fly this morning. Let's fix it now."
"Oh. Gosh. I don't have to explain it all again? Young man, that's the most civilised thing that has happened to me all week."
At 4am, the YB was asleep on a beanbag, travel pillow under his head, mumbling "kurang manis" every twenty minutes. Farid blamed the doctor from Part 4.
From Melbourne came one message: "Who are you people and what have you done with our queue?"
The envelope
At 4:30am, Aisyah, the youngest agent, brought Kavin a teh tarik. "If Saloma takes the easy calls, what happens to us?"
Kavin thought about Mrs Patterson. "She takes the calls that were burning us out, so we can do the ones that need us. Jessie's already planning new roles. Quality reviewers. AI supervisors."
"So we become Saloma's managers? With a raise?"
"Let's not get ahead of ourselves."
When she left, Kavin opened his bottom drawer, looked at the envelope, and tore it neatly in half.
5:47am
The storm had passed. The sale was a success. Then a call came in.
"Good morning, I'm Saloma, an AI assistant. How can I help?"
A smooth, confident voice replied: "Good morning, Saloma. I am also an AI assistant. And I was told to ask for you by name."
Kavin dropped his pen. Farid choked on his teh tarik. On the beanbag, the YB mumbled, "Next question."
An AI was calling an AI. And somebody had sent it.
To be continued in Part 6.
What this means for your BPO or contact centre
- Absorb peaks without long queues: sales events, launches and outages, with the AI handling routine calls in parallel.
- Keep running when staff can't get in: storms, floods or illness no longer mean a collapsing service level.
- Hand over with context: when a call needs a person, the agent sees a summary first, so customers never repeat themselves.
- Serve customers in their language: English across accents, Mandarin, Bahasa Malaysia, Tamil, Arabic and more.
- Ease burnout: agents spend their time on meaningful cases, with new roles in quality review and AI supervision.
Frequently asked questions
What is Saloma? Saloma is an enterprise AI voice agent by Astiostech, a Malaysian technology company. It handles inbound and outbound phone calls in natural speech in more than 20 languages, including Bahasa Malaysia, English, Mandarin, Tamil and Arabic. Saloma was formerly known as Vocalis.
How can a BPO use an AI voice agent? To answer high-volume, repetitive calls such as order tracking, returns and simple account questions, while human agents handle complex, sensitive or escalated cases.
Will AI voice agents replace call centre agents? They take over repetitive work so people can focus on cases that need judgement and empathy. Many contact centres also create new roles such as quality reviewers and AI supervisors.
What happens when a caller needs a human? Saloma transfers the call and passes on a summary of the conversation, so the customer doesn't have to start again.
Can an AI voice agent help during disruptions? Yes. Because it handles routine calls in parallel, service can continue when staff numbers drop suddenly, for example during floods.
Book a 15-minute demo for your contact centre
Read the practical guide: How BPOs and contact centres use AI voice agents to handle peak call volume.
This is a work of fiction. Wombat & Co, the BPO and every character are imaginary. Any resemblance to real people or organisations is coincidental. The goose remains unconfirmed.
Coming up in Part 6, the finale: Who sent the AI? Why does it want Saloma by name? And why are 300 more about to call?