钛媒体:引领未来商业与生活新知 04月16日
China's Deepexi Files for Hong Kong IPO Amid Enterprise AI Pivot
index_new5.html
../../../zaker_core/zaker_tpl_static/wap/tpl_guoji1.html

 

中国人工智能初创公司 Deepexi 计划在香港联交所 IPO,有望成为首家在港上市的企业级 AI 解决方案提供商。该公司成立于 2018 年,最初专注于企业“中间平台”领域,后来转型为全栈式企业 AI 应用提供商,推出了 FastData 和 FastAGI 等旗舰平台。尽管营收快速增长,但 Deepexi 仍处于亏损状态,主要归因于研发投入和股权激励。该公司计划利用 IPO 资金进一步发展平台、扩大模型平台,并在销售、交付和研发方面扩大团队,重点关注消费零售、医疗保健、制造和交通运输四个行业。

🚀 Deepexi 成立于 2018 年,最初专注于企业“中间平台”领域,旨在集中化企业数据和运营基础设施。近年来,公司迅速转型,成为一家全栈式企业 AI 应用提供商,推出了 FastData 和 FastAGI 等平台。

📈 在 2024 年,Deepexi 占据了中国企业级大模型 AI 应用市场 4.2% 的份额,预计 2025 年将增至 4.4%。FastData 贡献了 62.8% 的收入,FastAGI 贡献了 37.2%。

💰 尽管营收增长迅速,2024 年达到 2.43 亿元人民币,但 Deepexi 仍处于亏损状态,2024 年亏损额高达 12.5 亿元人民币,主要原因是研发投入和股权激励。过去三年,累计亏损超过 24 亿元人民币。

💡 Deepexi 计划利用 IPO 募集资金进一步发展 FastData Foil 集成平台,扩大 Deepexi 大模型平台,并构建内部计算基础设施。同时,公司计划扩大销售、交付和研发团队,重点关注消费零售、医疗保健、制造和交通运输四个行业。

⚠️ 招股说明书指出,Deepexi 面临重大业务风险,包括资金需求、技术创新压力和商业化挑战。但公司 CEO 赵杰辉认为,企业 AI 的未来在于执行力,将架构转化为应用。

(Image Source: Photo by Lin Zhijia, TMTPost AGI Editor)

AsianFin -- Chinese AI startup Deepexi Co., Ltd. has filed for an initial public offering on the Hong Kong Stock Exchange, aiming to become the first enterprise-level AI solutions provider to go public in the city in 2025.

The IPO application, submitted on Tuesday, lists CITIC Securities, Minsheng Capital, Guotai Junan International, Pu Yin International, and BOCOM International as joint sponsors.

Founded in 2018, Deepexi was initially recognized as a key player in the enterprise “middle platform” space—technology architecture designed to centralize data and operational infrastructure across business units. But in recent years, the company has pivoted sharply amid the generative AI boom, transforming into a full-stack enterprise AI application provider with flagship platforms such as FastData and FastAGI.

CEO Zhao Jiehui, a former Huawei and Alibaba Cloud executive, has been the architect of this transformation. Zhao, 45, spent over two decades in the IT sector and played a critical role in Alibaba's early middle platform architecture before launching Deepexi. Under his leadership, the company has positioned itself as a bridge between traditional enterprise systems and next-generation AI models.

The turning point came in 2023, during the global frenzy sparked by OpenAI's ChatGPT. Deepexi quickly launched FastAGI, a generative AI solution tailored for enterprise needs. By 2024, the platform had seen large-scale commercialization across verticals such as retail, manufacturing, and healthcare. According to the prospectus, Deepexi captured 4.2% of China's enterprise-level large model AI application market in 2024 and is expected to grow that share to 4.4% in 2025.

Despite strong growth, the company remains deeply unprofitable. Deepexi reported revenue of RMB 243 million in 2024, up from RMB 100 million in 2022—representing a compound annual growth rate of 55.5%. However, losses have widened substantially, reaching RMB 1.25 billion last year, over five times its annual revenue. The company attributes these losses to aggressive R&D investments and share-based payment expenses. In total, Deepexi has lost over RMB 2.4 billion in the past three years.

Deepexi's data intelligence solution, FastData, accounted for 62.8% of 2024 revenue, while FastAGI contributed the remaining 37.2%. The firm served 245 enterprise clients last year, up from 129 in 2022.

With nine funding rounds and backing from investors including Hillhouse Capital, IDG Capital, and the Shanghai AI Investment Fund, Deepexi was valued at $1 billion as early as 2021. Zhao remains the largest shareholder with 34.95% equity and 64.17% of voting rights, thanks to a dual-class share structure.

Looking ahead, the company plans to use IPO proceeds to further develop its FastData Foil integration platform, expand the Deepexi large model platform, and build in-house computing infrastructure. Deepexi also aims to grow its workforce in sales, delivery, and R&D, with a focus on four industries: consumer retail, healthcare, manufacturing, and transportation.

Still, the prospectus notes significant business risks, including capital demands, tech innovation pressures, and commercialization hurdles. But Zhao appears confident that the next wave of enterprise AI will be won not just by tech, but by execution—turning architecture into application.

 (Note: 1 USD equals about RMB 7.2)

更多精彩内容,关注钛媒体微信号(ID:taimeiti),或者下载钛媒体App

Fish AI Reader

Fish AI Reader

AI辅助创作,多种专业模板,深度分析,高质量内容生成。从观点提取到深度思考,FishAI为您提供全方位的创作支持。新版本引入自定义参数,让您的创作更加个性化和精准。

FishAI

FishAI

鱼阅,AI 时代的下一个智能信息助手,助你摆脱信息焦虑

联系邮箱 441953276@qq.com

相关标签

Deepexi AI IPO FastData FastAGI
相关文章