How a Korean immigrant’s English struggle helped deaf users make phone calls Joseph Lee, a Korean immigrant who struggled with English phone calls, founded InnoCaption in 2007 to provide real-time captions for deaf and hard-of-hearing users. The FCC-funded service now helps users make calls via mobile devices, using a hybrid of AI and human captioners. The company aims to expand similar tools to countries like Korea as aging populations increase demand for accessible communication. Even in the age of AI, captioned calls show why basic phone access still matters When Joseph Lee moved from Korea to the United States in the late 1980s, ordinary phone calls were among the hardest parts of daily life. Still learning English, Lee relied on closed captions to understand movies and television. But phone calls offered no such help. Before important calls, he would write down what he wanted to say and sometimes fax it in advance so he could follow the conversation more easily. The experience stayed with him. Years later, it became the basis for InnoCaption, a California-based company he founded in 2007 to provide real-time captions for phone calls, especially for people with hearing loss. The service now helps deaf and hard-of-hearing users in the US make and receive phone calls through captions on mobile devices. It is funded through the US Federal Communications Commission’s Telecommunications Relay Service program, which is designed to give people with hearing or speech disabilities telephone access comparable to that of hearing users. For many users, the problem is not that they cannot communicate at all; it is that phone calls remain difficult in situations where texting or email is not enough. “There are still many situations where phone calls are the only option,” said Paul Lee, InnoCaption’s chief executive officer. “Customer service, hospitals, government agencies, employment offices and first-round job interviews often still happen over the phone.” That makes captioned calling less a convenience than a matter of access. Many users are older adults with acquired hearing loss who still communicate through spoken language but need captions to catch words they miss, especially when the other person speaks quickly, has an accent or is in a noisy environment. InnoCaption initially relied on live stenographers to transcribe calls in real time. It later added automated speech recognition, allowing users to choose between AI captions and human captioners during a call. Paul Lee said the hybrid model remains important because AI still struggles in some real-life conditions. “AI has improved tremendously, but there are still situations involving heavy accents, unique speech patterns or background noise where people prefer a human captioner,” he said. The company has also moved into related tools, including text-to-speech functions and multilingual captioning. But its core service remains focused on making ordinary phone calls usable for people who might otherwise avoid them. For Joseph Lee, now founder and chairman, the idea still goes back to the frustration he felt before phone calls as a newcomer in the US. “I knew how much captions had helped me,” he said. “When I learned how many people with hearing loss could benefit from this technology, I couldn’t let the idea disappear.” The FCC-funded model is specific to the US, but Joseph and Paul Lee say similar tools could become more relevant in countries such as Korea, where aging populations are likely to increase demand for accessible communication. “As AI changes how people communicate,” Paul Lee said, “our goal remains the same as it was 10 years ago — using technology to help people connect.” jychoi@heraldcorp.com