Tomato.ai, a startup co-founded by CEO Ofer Ronen and CTO James Fan, has unveiled a zero-shot machine learning model designed to soften accents in real-time for call center agents. The company, which recently raised $10 million in funding, aims to improve communication and build trust between agents and customers, particularly in offshore call centers.
“Tomato.ai is all about making communication easier around the world by reducing the level of accent heard by the listener on the content of the conversation,” Ronen explained in an interview with VentureBeat. “The goal is, is twofold. One is to improve the ability to follow what’s being said intelligibility. And then the other one is to improve trust.”
The accent softening technology addresses a critical pain point for many offshore call centers in countries like India, the Philippines, and Pakistan. These centers often struggle to achieve the same performance metrics as their onshore counterparts, leading to lower customer satisfaction scores and difficulty securing new business.
Empowering human agents with AI
Ronen, who previously worked on contact center AI at Google, saw an opportunity to use AI to help human agents rather than simply automating their jobs. “We saw that there’s an opportunity to use AI to help humans and give them superpowers,” he said. “And that the voice AI piece was like, not quite yet offered at that scale. And that, you know, we were well equipped to do that having like that background.”
Tomato.ai’s solution is designed to be easily implemented, installing on lower-end PCs in just five minutes with no integration required. This frictionless onboarding was a key priority for Ronen, who wanted to avoid the complex integrations he often encountered at Google.
Zero-shot learning and ethical considerations
The technology also boasts zero-shot learning capabilities, meaning it can adapt to new speakers without any warm-up period. “As you start speaking, it just does the right thing without needing to warm up,” Ronen noted.
While the accent softening technology has the potential to significantly improve call center performance, it also raises ethical questions around cultural sensitivity and potential bias. Ronen emphasized that Tomato.ai is committed to maintaining speakers’ identities and offers different levels of accent softening to give users more control.
“We have different versions of our models that have different levels of softening,” he explained. “And so you could we believe in giving options to folks. And so if one model might work better for a use case, or a person, you know, they couldn’t have more of an American accent or more of a blend of their original and American.”
A game-changer for offshore call centers
From a business perspective, Tomato.ai’s technology could prove to be a game-changer for offshore call centers. By enabling agents to communicate more effectively with customers, these centers may be able to close the performance gap with their onshore competitors and secure more business.
The global call center market is expected to reach $496 billion by 2027, driven in part by the increasing adoption of AI and automation technologies. With its innovative accent softening solution, Tomato.ai is well-positioned to capture a significant share of this growing market.
However, the company will need to navigate the ethical complexities surrounding accent modification carefully. As AI continues to transform the call center industry, striking the right balance between efficiency and cultural sensitivity will be critical for long-term success.
The post Tomato.ai launches zero-shot accent softening model to revolutionize call center industry appeared first on Venture Beat.
