PT GoTo Gojek Tokopedia Tbk. (GOTO) is advancing its artificial intelligence (AI) initiative, Sahabat AI, with stronger support from Nvidia. Based on a report from Bisnis.com, GOTO’s CEO, Patrick Walujo, announced that Nvidia has expressed keen interest in furthering the development of this AI-powered platform.

“Sebenarnya, saya baru saja mendapat pesan dari Jensen kemarin bahwa NVIDIA kini ingin mendukung lebih jauh pengembangan dari Sahabat AI,” said Walujo during the Indonesia Pe-Vc Summit on Thursday, January 16, 2025. Jensen Huang, Nvidia’s CEO, personally communicated the company’s commitment to deepening their collaboration with Sahabat AI.

Sahabat AI, co-created by Indosat and Nvidia, represents a significant milestone in Indonesia’s AI ecosystem. Designed to process vast amounts of data in Bahasa Indonesia and various regional dialects, it aligns international expertise with local innovation to address Indonesia’s unique needs. The AI platform is integrated with Nvidia's full-stack AI solutions, including NVIDIA AI Enterprise and NeMo.

According to Walujo, the collaboration with Nvidia is pivotal for GOTO. “This partnership is crucial. We are gradually incorporating these initiatives into our daily operations,” he said. Nvidia’s global expertise in chip design and AI technologies has been instrumental in driving the platform’s capabilities.

When asked if Nvidia might invest in GOTO, Walujo replied, “Nvidia berinvestasi di Goto? Saya tidak tahu. Saya tidak tahu jawabannya. Namun, kami bekerja sama dengan mereka.” He emphasized that the relationship between GOTO and Nvidia remains centered on collaboration and technical integration rather than direct investment.

Sahabat AI serves as an open ecosystem, bringing together research institutions, universities, government bodies, and private sector partners. This initiative highlights Indonesia’s commitment to harnessing cutting-edge AI for local applications while preserving cultural and linguistic nuances. Supported by AI Singapore and Tech Mahindra, the platform is set to launch with large language models (LLMs) featuring 8 and 9 billion parameters.

SOURCE: BISNISTEKNO | PHOTO: INDOSAT