Flitto Signs MOU with Upstage for AI Language Data Building Endeavor

Collaboration aims to bolster competitiveness in the enterprise language model market through the collection and construction of low-resource language datasets in Asia.

Seoul, May 9, 2024: Flitto (CEO Simon Lee, KOSDAQ 300080), an AI data solution provider and integrated language platform, announced on the 9th that they have signed a memorandum of understanding (MOU) with Upstage, an AI company specializing in large language models, for the construction of AI language data.

The two companies will collaborate to collect and build datasets for low-resource languages in Asia, such as Thai, Japanese, Lao, and Khmer, with the goal of enhancing the performance of large language models (LLMs).

The agreement encompasses several key points. ▲Collaboration on constructing a benchmark dataset for Ko-LLM, the leaderboard for Korean LLMs. ▲Cooperation in operating a multilingual LLM leaderboard. ▲LLM localization through the establishment and utilization of low-resource language data. ▲Strengthening partnership through data supply for the establishment of small language models (sLLM) for business integration.

The strategic cooperation between the two companies, centered around language AI development, is expected to yield tangible results in the near future.

Through this partnership, Flitto aims to actively contribute to the advancement of language models by leveraging their language data scaling capacity and advanced AI technology. Their extensive expertise in language data construction, including building multilingual parallel corpora and providing copyright issue-free text, image, and speech datasets, is expected to give the AI data provider a solid competitive edge.

Upstage’s strategy is to secure high-quality, low-resource language data and to expand the language support of its pre-trained LLM Solar to develop customized models specialized for various regional languages, including Southeast Asia. Solar currently supports Korean and English, with plans to expand to Japanese and Thai by the end of the year.

Flitto CEO Simon Lee stated, “Training on low-resource languages has emerged as a key factor in the performance of large language models.” He added, “Through the partnership between the two companies, we aim to demonstrate how the synergy between high-quality AI data and advanced technologies can positively influence the domestic generative AI ecosystem.”

Upstage CEO Sung Kim emphasized, “Together with the generative AI boom driven by language models, securing quality language data is an essential task,” adding, “Through our partnership with Flitto, Upstage will endeavor to build advanced datasets, enabling a wider global audience to experience the innovation brought by generative AI technology.”

Photo 1. On the 8th, at Flitto's headquarters in Gangnam-gu, Seoul, Flitto CEO Simon Lee (third from the left) and Upstage Executive Vice President Soon-il Kwon (fourth from the left) pose for a commemorative photo with employees from both companies during the signing ceremony of the MOU for AI language data building endeavor.

* Additional information about the photo: (From left to right) Flitto Head of Product Strategy Group Ki-young Shin, Flitto Head of Communication Group Chung-won Park, Flitto CEO Simon Lee, Upstage Executive Vice President Soon-il Kwon, Upstage Principal Research Engineer Chan-jun Park, LLM Business Development and Sales Lead Min-sung Kim.