[AI Brief] Wrtn Selects IPO Underwriters, Eyes Leap as a Global AI Company
IT DAILY ·
✦ AI Summary
Wrtn Technologies announced on the 23rd that it has selected Mirae Asset Securities as the lead underwriter and NH Investment & Securities as the co-underwriter for its IPO.
Wrtn Technologies said it raised KRW 100 billion in Series C funding last month and was valued at more than KRW 1 trillion for the first time in Korea’s AI services sector.
AWS unveiled the open source AI agent 'Strands Harness' on the 23rd, and Enhance and GS Caltex announced that they will jointly carry out a project to develop an AI model based on manufacturing tacit knowledge.
Wrtn Technologies announced on the 23rd that it has selected its IPO underwriters. Mirae Asset Securities will serve as the lead underwriter for the IPO, and NH Investment & Securities will be the co-underwriter.
Last month, Wrtn Technologies raised KRW 100 billion in Series C funding. Through this, it became the first company in Korea’s AI services sector to be valued at more than KRW 1 trillion.
As a result, Wrtn Technologies’ cumulative funding raised reached about KRW 230 billion. Earlier this month, Wrtn Technologies officially announced the start of its public listing process.
Since then, Wrtn Technologies has been working to select its underwriters, and through this announcement it has finalized its lead and co-underwriters.
Wrtn aims to file for preliminary review next year and plans to coordinate details with its underwriters going forward. Kim Jin-tae, head of the IPO division at Mirae Asset Securities, said Wrtn has shown explosive growth in the AI services sector and proved its unmatched market competitiveness. He also said Wrtn is showing J-curve growth momentum and expressed his commitment to supporting a successful listing.
Lee Se-young, CEO of Wrtn, said the company’s overseas revenue from its global service is higher than its domestic revenue. He also said Wrtn’s competitiveness has been proven in the global market.
Lee said the company’s goal is to leap forward as a global AI company through a capital market listing. He added that Wrtn is currently positioned as Korea’s leading AI startup.
In separate news, AWS unveiled an open source AI agent called Strands Harness on the 23rd.
AWS said Strands Harness can reduce token costs by 28%.
AWS cited the high barrier to entry in the agent development process as the background for unveiling Strands Harness. According to AWS, developers had been evaluating existing tools such as Claude Code and Codex, with one of their strengths being that they work without additional setup. By contrast, building an agent directly required connecting multiple basic components one by one, making it difficult to deliver the same user experience, AWS said.
Strands Harness is an execution environment that assembles components for running AI agents. It provides the process in a preconfigured form, can be run immediately on a developer’s local PC, and can be deployed to the cloud of choice. It also supports running agents with the AI model selected by the developer and a single line of code.
Strands Harness is offered under the Apache 2.0 license and can also be modified. Its design goal is not to be limited to specific coding tasks, but to serve as a general-purpose agent usable across a range of work.
According to AWS’s own tests, under the same conditions using Claude or GPT models, Strands Harness reduced average token costs across six benchmarks by 28%. Its accuracy was assessed as equal to or better than that of existing tools such as Claude Code and Codex. In tests using the 'Fable 5' model, it cut costs by 77% compared with Claude Code, and its 'Terminal Bench 2.1' score also outperformed Claude Code.
Enhance announced on the 23rd that it and GS Caltex will jointly carry out a project to develop an AI model based on manufacturing tacit knowledge. The project aims to develop an AI model that implements the know-how of experienced engineers' experience and judgment.
The two companies plan to link equipment data, maintenance history, and technical documents through ontology. The idea is to structure expert knowledge by connecting various data and documents accumulated at manufacturing sites.
They also plan to build an AI chatbot that presents abnormality causes, inspection steps, and decision grounds. Through this, they aim to use the manufacturing tacit knowledge-based AI model in chatbot form.
Enhance will carry out the research and development. The R&D will link GS Caltex’s structured equipment data, unstructured documents, and maintenance history, and further connect judgment criteria and exception-handling experience by securing interviews with field engineers. The source data includes 832,996 GS Caltex pieces of equipment and about 2,900 unstructured documents. The relationships to be structured are equipment, parts, abnormal signs, causes, and actions.
Based on the knowledge built in this way, Enhance will implement a field-use 'Guide chatbot.' Field staff will be able to ask natural-language questions about equipment abnormalities and maintenance situations. The chatbot will present possible causes, inspection steps, past maintenance records, and technical standards. It will also be developed so that users can check original reference text for judgment support.
Lee Seung-hyun, CEO of Enhance, said the company plans to build a system that supports field staff in checking the information they need and the grounds for judgment, while helping them solve problems more efficiently.
Source: IT DAILY · Yang Seung-gap
Original: https://www.itdaily.kr/news/articleView.html?idxno=241824
References
This article was produced with the help of an automated content generation algorithm.
Source: IT DAILY
View originalThis article was summarized and organized by BizCrush based on the original article from IT DAILY. For exact quotations and full details, please refer to the original article.