Ji-Yan Wu

L

ocWorld isn’t just a venue for language professionals to network and have fun. It’s also an incubation chamber for new ideas. From the informal conversations to the scheduled sessions and the booth demonstrations, the entire conference is a hub for inspiration and learning.

Perhaps no LocWorld institution, however, is more focused on progress than the Process Innovation Challenge (PIC). A competition among industry peers to deliver the most forward-thinking, outside-the-box ideas, the PIC pits some of the industry’s brightest minds in a head-to-head battle to introduce new efficiency, productivity, quality, and sustainability into language work.

The 18th PIC, held June 4 in Malmö, Sweden, was no exception. After some difficult evaluations, the panel of judges came to a decision: Ji-Yan Wu of Centific took the top prize for his presentation, “Smarter Video Localization With AI: Adapting On-Screen Content for Global Audiences.” MultiLingual reached out to Wu for his thoughts on the victory and further insights into his award-winning innovation.

Congratulations on winning the latest PIC! Could you briefly summarize your innovation?

We developed a platform called Flow that will localize videos using agentic artificial intelligence (AI) in just one or two clicks. The agentic AI doing the localization understands video context, so it translates meaning, not just words. It automatically generates natural-sounding voiceovers, replaces on-screen text such as subtitles and signs, and synchronizes everything — voice, video, and visual elements — seamlessly.

What was your inspiration, and how did you balance this project with your other responsibilities?

We looked at the industry and saw that the current wave of AI-powered localization was being driven by the private sector to localize products that can be sold to consumers for profit. And we thought there was something missing. If AI is as transformative as people say it is, then it has to serve everyone. We thought we could do better. We looked at the high cost of video localization — you need a team of localization specialists, software tools like Adobe, a professional recording studio to record voiceovers, and voice talent. This all takes resources that someone like a teacher who serves underrepresented communities may not have.

Developing the platform and keeping up with my work wasn’t easy. A lot of late nights and weekends, but we were driven by the idea that we could do it. It wasn’t fully featured like it is today. It was a simple web app at its beginnings. But as more internal team members started to see the vision, they signed on to help. Then it became more stable, faster, and easier to navigate. There’s always room for improvement, but what you see today is the result of lots of internal team members’ hard work, and a lot of coffee.

What sort of operation stands to benefit the most from this platform?

We think that e-learning companies can benefit the most from a platform like Flow. They have volumes of educational video content that is already being localized today for the most popular language groups. These videos have multiple challenges to overcome.

For example, subtitling them becomes problematic when the presenter in the video starts showing slides. Subtitling the words of the speaker and the content on the slides would create a very bad learning experience. I like visiting the Great Wall of China — but I don’t want to read a wall of Chinese text when learning something new! If we took away the subtitles because the speaker was speaking in a human voice and the slides were automatically translated in the video, I could learn more naturally, like the teacher was teaching the content for me. Now put that power in the hands of the educators who want to share their knowledge with all — not just people of major language groups, but all language groups, even the underrepresented ones.

How did it feel to take the top PIC prize?

To be honest, I didn’t think we would win. There are so many great ideas and talented people who have worked in this field for so long. When I heard my name called and realized Centific had won, I was stunned. I thought they were mistaken. And I think my teammates were equally surprised. When we found out together, I heard them scream from the front row. I think they screamed in surprise more than anything else. I was proud to represent a whole team of people in China, Europe, and the United States. I was standing on the stage, but I felt like I was standing on the shoulders of some hard-working folks on the team.

Does this win fuel your motivation to tackle other persistent challenges? Are there any other tricky problems of particular interest?

We don’t have any concrete plans right now, but we have a focus. The market is so concerned about how to make money and save money with AI. And in some sense, they are justified. But I think this is the wrong focus that just leads to more boring ideas. The real challenge, the more exciting challenge, is to ask ourselves, “How can we use AI for the good of all? How can AI lift up those who cannot stand on their own? How can AI level the playing field?” I don’t know exactly where this challenge will take us, but I’m certain that if we focus on this, we will come up with something great.

Is there anything else you want to add?

I’d like to thank my manager, Wei Zhang, who permitted me to work on this project and dedicated his own time, sweat, and tears (there were many) to this project. But we are not done yet. So I’d like to thank my colleagues, especially David No and Ge Zhu, who also make important contributions to our work. Thank you to my company, Centific, which has created a work culture that allows for innovation to take place. Thank you.