KBRN, Bandung: The National Research and Innovation Agency (BRIN), through its Center for Data and Information Science Research (PRSDI), has developed an artificial intelligence (AI)-based Large Language Model (LLM) alignment tool to help preserve endangered local languages in Indonesia.

In collaboration with the Faculty of Cultural Sciences at Hasanuddin University and the Faculty of Computer Science at Handayani University, BRIN has built a multilingual translation platform powered by LLM.

The platform enables text translation, audio-to-text and video-to-text conversion, and automated subtitle generation in multiple languages.

The prototype, currently under development, uses Bugis as the base model. The Bugis language—also known as Basa Ugi—is spoken by the Bugis ethnic group in South Sulawesi.

It is one of the largest regional languages in Indonesia. Bugis has a rich oral tradition and its own ancient script, Lontara. Despite its cultural significance, Bugis is considered at risk due to a generational shift toward Indonesian and foreign languages.

"With a single Bugis translation model, it can translate into various foreign languages such as English, Spanish, and Russian—and vice versa," said Yuyun, a senior researcher at PRSDI, in a statement on Saturday, June 28, 2025, as quoted by brin.go.id.

The platform, accessible at http://nlp-prototype.brin.go.id/ (via local network), currently incorporates approximately 14,000 Bugis words. Yuyun hopes to expand the platform in the future by including other regional languages to widen its preservation impact.

However, the platform still faces limitations, especially in terms of building a comprehensive vocabulary. "A lack of data and difficulty in finding qualified local language experts pose significant challenges in developing an accurate dataset," Yuyun explained.

He emphasized that collecting local language vocabulary is essential for effective translation between Indonesian and regional languages—an important factor in ensuring that the technology can genuinely contribute to language preservation.

Yuyun hopes this platform will enhance understanding of local languages in the digital era. He added that the initiative also supports digital transformation rooted in Indonesian local wisdom.

Through this innovation, BRIN aims to safeguard cultural diversity with the help of modern technology. The project represents a strategic step toward integrating regional languages into Indonesia's evolving digital ecosystem. ***