Case Study: FM Wakayama achieves continuous, broadcast-quality AI announcements for disaster response with Amazon Polly

FM Wakayama - Customer Case Study

FM Wakayama, the nonprofit operator of Banana FM serving about 500,000 people around Wakayama City, needed a way to deliver timely news and weather around the clock without keeping on‑call staff during nights and disasters. To solve this, the station adopted Amazon Polly’s text‑to‑speech service to create Nanako, an AI announcer that could produce human‑level, broadcast‑quality audio for emergency and regular programming.

By integrating Amazon Polly with its Ontime Player and Da Capo systems (and securing access via Amazon Cognito), FM Wakayama built Nanako in about a month and went live in July 2017; the service costs only ¥400–800 per year and was far cheaper than alternatives. During Typhoon 18 the AI announcer provided five hours of continuously updated information—while up to 4,400 households were without power—and FM Wakayama was one of the only broadcasters able to maintain live updates, reducing the need for on‑call staff and making Amazon Polly a reliable part of its disaster response.

FM Wakayama

Seiji Yamaguchi

Cross-Media Station Director