[IT Donga x Spark Labs] The Southeast ICT Innovation Square Expansion Project, hosted by the Ministry of Science and ICT and the National IT Industry Promotion Agency (NIPA), and organized by the Busan Information Industry Promotion Agency, aims to nurture digital talent in the Busan, Ulsan, and Gyeongnam regions, drive digital transformation in local industries, and foster a startup ecosystem. In this project, IT Donga introduces startups incubated by Spark Labs.
K-content is growing rapidly enough to succeed in the global market. Over the past decade, the domestic web novel market has expanded nearly 100-fold, from tens of billions to over 1 trillion KRW. However, despite the growing power of the content, very few web novels actually make it overseas. Unlike video content, text-based content has a structure where translation costs far exceed production costs. Moreover, the process of exporting content overseas — beyond simple translation — is extremely difficult for small and medium-sized publishers.
HiveMind (CEO Warren Kim), the company behind the global content publishing platform Aurorah, is a startup working to solve this problem using artificial intelligence (AI) technology. Aurorah focuses on publishing domestic content, especially web novels, to global markets. What sets it apart is that it supports the entire publishing process — from translation to review, editing, distribution, and settlement. HiveMind aims to turn Aurorah into a global story IP hub and establish a new standard for K-content global distribution.
Warren Kim, CEO of HiveMind, is a former accelerator with nearly 10 years of experience investing in content startups in animation, metaverse, and webtoons. Witnessing many publishers struggling with global expansion inspired him to found the company. “For webtoons, dramas, or films, production budgets reach tens of billions, so translation costs are relatively minor. However, web novels have very low production costs, but translation costs around 70–80 KRW per character, often resulting in tens of millions of KRW in losses per title,” he explained.
He also discovered that publishers who managed to complete translation often faced even greater difficulties with distribution and settlement. With over 40 domestic platforms having different settlement formats, going overseas adds further complications such as exchange rates and local entity requirements.
To address this, HiveMind built a platform that integrates all five stages of publishing. “Translation is ultimately just a tool to reach overseas readers and generate revenue,” said CEO Warren Kim. “If review, format conversion, distribution, and settlement are not handled properly, it’s all meaningless.”
HiveMind acts as an overseas publisher, allowing domestic publishers and authors to easily present their works to foreign markets just as they would domestically. The business model is designed to minimize the burden on content providers. Publishers and authors do not have to bear upfront translation and localization costs; instead, they share future IP revenue equally. This structure removes the initial investment barrier and also improves previously opaque overseas settlement practices.
CEO Warren Kim believes that as AI translation technology becomes more standardized, quality alone is no longer enough for differentiation. Therefore, the company emphasizes specialized K-content training data and human review collaboration as its key strengths.
HiveMind has secured diverse web novel and webtoon data through its partnership with StorinLab, which has a network of over 430 publishers. This allows the AI to maintain unique character personalities and preserve long-form narrative context, significantly improving translation quality.
After the AI produces the first draft, professional human translators conduct final review. HiveMind treats translators as co-creators of localized content, emphasizing that cultural and emotional adaptation (transcreation) is more important than pure linguistic skill.
Founded in 2023, HiveMind is now accelerating its global expansion roadmap. The Aurorah platform currently offers translation and file editing functions, with distribution automation and settlement systems scheduled for completion this year. The company plans to officially launch Aurorah as a SaaS service in July. In the second half of the year, it will focus on building foundations for global market entry.
In response to Amazon and other platforms tightening regulations against low-quality AI-generated books, HiveMind plans to register as an official publisher on major global platforms such as Amazon, Google, and Apple. The company has already begun supplying content to five platforms, including Webnovel (operated by Tencent’s China Literature), where popular works like Solo Leveling are officially licensed in English.
This year, HiveMind aims to validate demand for localized Korean content, accumulate export cases, and focus particularly on the Japanese market — the largest consumer and a strong content powerhouse — to establish Aurorah as a global story IP hub. CEO Warren Kim stated, “We plan to prioritize publishing original web novels that have succeeded as webtoons or dramas overseas but have not yet been translated.” He noted that many overseas readers who discover K-dramas or webtoons search for the original novels, but they are often unavailable in translation.
HiveMind received early support through Spark Labs’ incubation program as part of the Busan ICT Innovation Square Expansion Project (Southeast Region). CEO Kim said, “We will receive practical help including company diagnosis, mentoring, investment consulting, PR, and networking this year,” adding that the company aims to validate its business potential with domestic and international investors and create opportunities for global market entry. Finally, he emphasized, “HiveMind’s vision is to create a place where all stories gather and spread across the world. Our goal is to overcome barriers in translation and distribution with AI and establish Aurorah as the new global standard platform for K-content overseas distribution.”