US company creates solution to correct the way Indian call centre agents sound to American customers
Sanas, which claims to be at the forefront of AI innovations that blur the line between machine-generated enhancements and human interactions, claimed that its product has been developed with the goal of “reducing accent-based discrimination”.
Teleperformance, the world’s biggest call centre operator is using artificial intelligence (AI) to ‘”neutralise” Indian accents for Western customers, according to a report in The Telegraph. The company said it was applying real-time AI software on phone calls to remove the accents of English-speaking Indian customer service agents that will enhance clarity and improve customer interactions.
The French company has around 90,000 employees in India and tens of thousands in other countries which serve customers in the UK including the government, the NHS, Vodafone and eBay. The AI-driven solution has been developed by Palo Alto-based startup Sanas which also includes background noise cancellation technology that filters out disruptive sounds such as office chatter, sirens, or even crowing roosters to improve call quality.
“It’s a technology that allows [us] to neutralise accents in real time without any data storage,” Teleperformance’s Markus Schmitt told investors.
“You have obviously the issue we talked about human connection, human empathy. We have first implemented Sanas with clients in India. And sometimes, there is a difficulty people in India talking and vice versa with clients from the US.”
Sanas, which claims to be at the forefront of AI innovations that blur the line between machine-generated enhancements and human interactions, claimed that its product has been developed with the goal of “reducing accent-based discrimination”.
Notably, Teleperformance has taken a stake in the US-based company which will see Teleperformance’s calls used to train the company’s accent software.
“In a world that is ubiquitous with AI, the element of the human … will be also equally important because it’s about building human connection and having this element of human empathy, connectivity will be something that will be equally valuable in the future,” said Teleperformance’s chief executive Thomas Mackenbrock.
Social media reacts
As the news went viral, social media users had varied reactions to the implementation of the accent-changing AI-software.
“I don’t mind Indian accents. I’m just glad to be talking to a real person who wants to help, and usually can,” said one user, while another added: “I don’t know if it solves any problem. I just don’t want to be scammed.”
A third commented: “I’d honestly like to talk to the real person with the real accent.”
Amid the rise in AI chatbots that can handle several customer queries simultaneously, call centre companies have been under pressure to evolve.
Chat with GPT 4.5 now
GPT-4.5 was designed to be a better conversation partner, making discussions feel warmer and more intuitive. In tests, human evaluators rated it higher than GPT-4o in almost every category.
OpenAI has launched GPT-4.5, its newest and largest AI language model. At first, it will be available as a research preview for ChatGPT Pro users. OpenAI is calling it their “most knowledgeable model yet,” but warns that it may not perform as well as some other models like o1 or o3-mini. In a post on X (formerly Twitter), OpenAI CEO Sam Altman acknowledged that GPT-4.5 is a “giant, expensive model” and that it “won’t crush benchmarks.”
What’s new
The company says that GPT-4.5 comes with better writing skills, improved general knowledge, and a more refined personality than past versions. OpenAI claims conversations with GPT-4.5 will feel more natural, and it is better at recognising patterns and making connections.
However, OpenAI notes that GPT-4.5 is not a major breakthrough model. A leaked document revealed that it improves GPT-4’s efficiency by more than 10 times but does not introduce seven completely new reasoning capabilities. This mention was later removed from OpenAI’s official announcement.
It was earlier reported that OpenAI used its o1 model, codenamed Strawberry, to train GPT-4.5 with synthetic data. OpenAI confirms that GPT-4.5 was trained using a mix of new supervision techniques along with traditional methods like supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF), similar to GPT-4o.
Despite its limitations, OpenAI says GPT-4.5 has fewer hallucinations than GPT-4o and slightly fewer than o1. Researcher Raphael Gontijo Lopes stated that GPT-4.5 was designed to be a better conversation partner, making discussions feel warmer and more intuitive. In tests, human evaluators rated it higher than GPT-4o in almost every category.
GPT 4.5 is currently available for ChatGPT Pro users. OpenAI will expand GPT-4.5 access to Plus and Team users soon, followed by Enterprise and Edu users. It is also available now on Microsoft’s Azure AI Foundry platform, alongside models from Stability, Cohere, and Microsoft.
