Teleperformance SE, the world’s largest call center operator, has introduced AI-powered accent translation to soften the accents of English-speaking Indian workers in real time. The company claims the technology will improve customer experience by making agents easier to understand.
AI-Powered Accent Neutralization
The accent translation technology, developed by Palo Alto-based startup Sanas, also includes background noise cancellation to enhance call clarity. Teleperformance has invested $13 million in Sanas and is now its exclusive reseller.
Deputy CEO Thomas Mackenbrock stated that the tool “neutralizes the accent with zero latency”, increasing customer satisfaction and reducing call handling time. However, the company has not disclosed which clients are using the technology.
AI Investments and Market Response
As part of its broader AI strategy, Teleperformance plans to invest €100 million ($104 million) in AI partnerships this year. The company is integrating AI across its operations, using it for call transcriptions, employee coaching, and quality control.
Despite these advancements, Teleperformance shares fell over 10% in Paris after a disappointing earnings report. Analysts expressed concerns over weaker-than-expected margins, despite projected 3-5% revenue growth.
Concerns Over AI in Call Centers
While Teleperformance says AI will enhance rather than replace human agents, the rise of AI chatbots has sparked concerns about the future of human call centers. Last year, Klarna Bank AB claimed its AI assistant could handle the work of 700 full-time agents, sending Teleperformance shares to their lowest level since 2016.
Sanas says its technology aims to reduce accent-based discrimination, but some fear it may erase cultural identities and impact call centers in regions like the Philippines, known for their high-quality English-speaking workforce.
Despite these concerns, Mackenbrock emphasized that while AI will be everywhere, human connection remains crucial in customer service. Teleperformance is also expanding its role in AI model development, providing training and data labeling services for companies like OpenAI.
