12月15日，飞猪在2025年度假全球合作伙伴大会上宣布，将全面升级度假商家评分体系，为非标品 属性明显的度假商家和商品打造一套更科学、更透明的长期信用机制。同时，飞猪宣布将上线4款服务 于商家运营的AI智能体，以进一步降低度假商家的运营成本。 "开放和信任是度假行业发展的两大关键词。"飞猪CEO庄卓然分析，开放的平台模式比标准化的自 营模式更能激发供给侧活力，提升消费者选择的丰富度；一套强大的信任机制则能有效促进供需匹配， 让消费者始终敢决策、爱推荐，让好服务、好商家不断脱颖而出。在这两大力量的驱动下，度假生态就 能以健康、可持续的方式向前发展。他表示，作为在线旅游市场唯一的OTP，飞猪负责升级消费者和与 商家之间的信任体系、为商家打造更好用的生产工具；而留给商家最重要的命题，则是创造更多的新供 给。 和执行效率。以AI优化助手为例，当商家希望对已经发布的商品进行批量优化时，AI能够基于真实数 据主动分析、提供落地建议，并支持一键执行。 供给侧活力充沛，需求侧的增量空间就能发挥出更大价值。在庄卓然看来，新的一年平台上的度假 商家将迎来阿里大消费平台战略加持、AI全新生产力和入境游加速三大增长机遇。 今年6月 ...

本报讯 （记者梁傲男）12月15日，飞猪在2025年度假全球合作伙伴大会上宣布，将全面升级度假商家 评分体系，为非标品属性明显的度假商家和商品打造一套更科学、更透明的长期信用机制。同时，飞猪 宣布将上线4款服务于商家运营的AI智能体，以进一步降低度假商家的运营成本。 据飞猪透露，新一代度假商家评分体系已初步设计完成，即将进入内测阶段。该评分体系将围绕用户口 碑、产品履约、服务保障三大指标构建商家的"行为+信用"机制。与现行体系相比，新评分体系的指标 中产品视角和用户视角的评价维度更加丰富和多元。例如，在产品履约指标下还细分了行程合理性、产 品一致性、司导服务水平等原子化维度。另外，新评分体系还着重增加了经营维度数据，如累计服务人 次、88VIP好评率等。针对虚假好评、恶意差评等行业顽疾，新评分体系纳入了行为可信度等考量维 度，并优化商家的申诉机制。飞猪方面表示，绝不让在平台上用心经营、尽心服务的好商家受到不公正 因素的干扰。 综合来看，这套评价标准视角更宽、指标更细，能有效剔除噪声数据干扰，呈现出更加科学、客观的评 价结果。这既能让消费者在低频的休闲度假消费场景中敢于决策，同时也能在商家后台反向作用于问题 诊 ...

Investment Rating - The report suggests a positive investment outlook for the AI and financial services sector, highlighting the potential for significant advancements and cost reductions due to the release of DeepSeek R1 in 2025, which is expected to be a turning point for localized AI deployment in financial institutions [7]. Core Insights - AI applications are rapidly penetrating core business areas and back-office functions of various financial institutions, enhancing both internal efficiency and external value [1][7]. - The report identifies that most financial institutions are currently in the exploration and accumulation phase of AI application, with deep application being an inevitable trend [14]. - AI is expected to transform financial business processes and organizational structures, ushering in a new era of digital intelligence in finance [7]. Summary by Sections Investment Recommendations - The report recommends focusing on several sectors within the financial industry, including: 1. Financial information services with key stocks like Tonghuashun, Jiufang Zhitu Holdings, and Guiding Compass [8]. 2. Third-party payment services, recommending stocks such as Newland and Newguodu, with related stocks like Lakala [9]. 3. Banking IT, with recommended stocks including Yuxin Technology, Jingbeifang, and Guodian Yuntong [9]. 4. Securities IT, recommending stocks like Hengsheng Electronics and Jinzhen Shares [10]. 5. Insurance IT, with recommended stocks including Xinzhi Software and Zhongke Software [11]. Application Stages - Financial institutions' AI applications are categorized into three stages: 1. Initial exploration of large model applications. 2. Development of certain model application capabilities with data accumulation. 3. Achieving deep application of large models [14]. Application Value - AI applications provide value through: 1. Internal cost reduction and efficiency improvement, optimizing operational management and core business processes [21]. 2. External value extraction, enhancing marketing and customer service to improve sales conversion and customer value [21]. Application Pathways - Different types of financial institutions exhibit varied pathways for AI application deployment: 1. Large institutions leverage strong self-research capabilities for deep AI application penetration. 2. Smaller institutions focus on cost-effective solutions, utilizing lightweight models and integrated systems for agile development [26]. AI Empowerment in Banking - AI is enhancing front-office quality and efficiency, optimizing back-office processes across various banking functions [43]. - In credit risk management, AI models can analyze financial data to identify potential risks and improve decision-making processes [47]. AI Empowerment in Securities - The number of securities firms exploring large models is rapidly increasing, with applications extending across various business functions, including investment advisory and research [58][59].

Core Viewpoint - Beijing Haixue Education Technology Co., Ltd. (Haixue) has submitted an IPO application to the Hong Kong Stock Exchange, aiming to raise funds for AI research, business expansion, and acquisitions despite ongoing losses and increasing refund rates [2][6]. Financial Performance - Haixue reported revenues of 400 million yuan, 472 million yuan, 511 million yuan, and 232 million yuan for the years 2022, 2023, 2024, and the first half of 2025, respectively [3] - The net losses for the same periods were 186 million yuan, 175 million yuan, 90.7 million yuan, and 158 million yuan, with gross margins of 78.6%, 81.1%, 81.6%, and 85.8% [3] - The first half of 2025 saw a 125.7% year-on-year increase in losses, surpassing the total net loss for 2024, while revenue decreased by 5.3% year-on-year [3] Customer Acquisition and Retention - Haixue's customer acquisition costs increased, with figures of 414.8 yuan, 301.6 yuan, 428.7 yuan, and 441.2 yuan during the reporting periods [3] - The number of paid users fluctuated, with 687,400, 1,045,600, 770,400, and 333,100 users in the respective years [4] - Refund rates rose significantly, reaching 9.1%, 9.9%, 10.5%, and 15.9% during the reporting periods, indicating challenges in customer retention [4][5] Business Structure and Market Position - Haixue's revenue is heavily reliant on construction-related courses, which accounted for 68.2%, 61.5%, 59.3%, and 59.8% of total revenue during the reporting periods [6] - The company is attempting to diversify into new fields such as finance, law, and healthcare, but acknowledges that returns from these new ventures may initially be lower than from existing offerings [6] AI Technology Initiatives - In 2025, Haixue established an independent AI innovation team to develop AI applications for educational support, teaching assistance, and marketing [7] - R&D expenditures were significantly lower than marketing costs, with figures of 24.5 million yuan, 20 million yuan, 23.7 million yuan, and 9.2 million yuan during the reporting periods [7] - The company is facing negative operating cash flow, with a reported negative cash flow of 1.5 million yuan in the first half of 2025, raising concerns about its ability to sustain operations without additional financing [7]

Core Viewpoint - The company, HiLearn, has submitted its application for a mainboard listing on the Hong Kong Stock Exchange, aiming to capitalize on the growing online vocational training market while facing significant financial challenges and operational risks [1][15]. Company Overview - HiLearn, established in 2010, provides online vocational training services targeting both individual and corporate users, with courses in various fields including construction, emergency safety, finance, law, and healthcare [1]. - The company operates under three main brands: "HiLearn," "Jingjin," and "Shupeitong," covering basic training, advanced improvement, and corporate services [1]. Market Position - According to a report by Frost & Sullivan, HiLearn is the fourth largest online vocational training provider in China by revenue as of 2024, leading in construction vocational training and ranking second in emergency safety training [3]. - As of mid-2025, the company has approximately 4.3 million cumulative paid individual users and over 130 corporate clients [3]. Financial Performance - The company's revenue has shown stable growth, with figures of 400 million yuan, 472 million yuan, and 511 million yuan for the years 2022, 2023, and 2024 respectively, reflecting a compound annual growth rate of 13.0% [4]. - Notably, the gross profit margin has increased from 78.6% in 2022 to 81.6% in 2024, and further to 85.8% in the first half of 2025, significantly above the industry average of 60%-70% [4]. Profitability Challenges - Despite high gross margins, the company has reported substantial losses, with attributable losses of approximately 186 million yuan, 175 million yuan, 90.69 million yuan, and 158 million yuan for the years 2022, 2023, 2024, and the first half of 2025 respectively [5]. - The high marketing and sales expenses, which exceeded 60% of revenue, are primarily driven by a heavy reliance on human marketing teams, with employee costs constituting about 60% of marketing expenses [5][6]. Customer Acquisition Costs - The cost of acquiring a single paid user has fluctuated, with figures of 414.8 yuan, 301.6 yuan, 428.7 yuan, and 441.2 yuan from 2022 to the first half of 2025 [6]. - The one-time nature of vocational training services leads to a high turnover of users post-qualification, necessitating continuous investment in acquiring new users [6]. Business Structure Risks - The company's revenue is heavily dependent on construction-related vocational qualification courses, which consistently account for around 60% of total revenue from 2022 to mid-2025 [7]. - The "Shupeitong" brand, aimed at corporate services, remains underdeveloped, generating only 4.12 million yuan in revenue in the first half of 2025, contrasting sharply with the 98.2% revenue share from individual users [7]. Policy Dependency - The company's focus on policy-driven sectors like construction and emergency safety has provided short-term benefits but limits long-term growth potential due to its concentrated business model [8]. AI Technology Initiatives - HiLearn is exploring AI technology applications, having developed over 30 AI agents since 2024, including tools like AI Study Assistant and AI Academic Assistant [9][10]. - The AI Study Assistant has reportedly been used over one million times, leading to a 32.3% decrease in human query responses in Q2 2025 [10]. Research and Development Concerns - R&D expenditures from 2022 to 2024 were 24.5 million yuan, 20 million yuan, and 23.7 million yuan, representing less than 5% of revenue, with a 23.3% year-on-year decline in the first half of 2025 [11]. User Experience Issues - The company's refund rate has increased from 9.1% in 2022 to 15.9% in the first half of 2025, with numerous complaints regarding refund difficulties and service discrepancies [12]. Legal Risks - HiLearn is involved in 905 legal cases, with nearly 90% of these cases being education training contract disputes, indicating significant legal exposure [13]. Industry Outlook - The online vocational certification exam market in China is projected to grow from 34.5 billion yuan in 2024 to 44 billion yuan by 2029, with the online vocational training market expected to reach 60.2 billion yuan by 2029, reflecting a compound annual growth rate of 5.1% [14]. Future Plans - The company plans to use the funds raised from its IPO to expand product services, sales networks, and advance AI technology development, with a strategic focus on infrastructure, data collection, and AI application development over the next five years [15].

Core Insights - The Hong Kong IPO market remains active, with total fundraising reaching HKD 182.9 billion in the first three quarters of 2025, a significant increase of approximately 230% compared to the same period in 2024, and the number of IPOs rising by about 50% to 67 [1] - The company, HiLearn Education Technology Co., Ltd. (嗨学网), has submitted its listing application to the Hong Kong Stock Exchange, marking another significant IPO case in the vocational education sector [1] Company Overview - HiLearn is the fourth largest online vocational training service provider in China, having served 4.3 million paid learners [2] - The company operates under three main brands: "HiLearn," "Jingjin," and "Shupeitong," covering various vocational fields such as construction, emergency safety, finance, law, and healthcare [2] - Despite steady revenue growth, the company remains in a loss-making position, with revenues of RMB 400 million, RMB 472 million, and RMB 511 million from 2022 to 2024, and a loss of RMB 1.58 billion in the first half of 2025 [2] Financial Performance - The company's gross margin has shown a steady improvement, increasing from 78.6% in 2022 to 85.8% in the first half of 2025, indicating operational efficiency and content optimization [2] - The revenue structure is heavily reliant on consumer services, with 98.2% of revenue coming from individual users in the first half of 2025 [4] - The company has served over 130 enterprises and institutions but generated only RMB 4.12 million from its enterprise service brand "Shupeitong," indicating a need for growth in the B2B segment [4] Product and Service Offerings - HiLearn offers two main product categories: "HiLearn Classroom" for recorded and live courses, and "Jingjin Classroom" focused on efficient exam preparation with personalized learning plans [3] - The company has made strides in AI applications, launching over 30 AI smart agents to enhance learning and teaching processes, with significant user engagement reported [6] Market Trends and Opportunities - The vocational education sector is experiencing growth due to government policies aimed at improving quality and expanding access, with vocational undergraduate enrollment increasing from 41,400 in 2021 to nearly 500,000 in 2025 [8] - The demand for skilled talent is projected to grow, with a predicted talent gap of 30 million in key manufacturing sectors by 2025 [8] Competitive Landscape - HiLearn ranks fourth in China's online vocational training market and holds the top position in construction vocational training [10] - The company’s revenue structure is still heavily dependent on construction-related qualification courses, which accounted for around 60% of revenue from 2022 to the first half of 2025 [10] - The company faces challenges in diversifying its course offerings and expanding into non-construction fields to ensure sustainable growth [10]

Core Insights - The company reported a revenue of 920 million yuan for H1 2025, a 17.7% increase year-on-year, with a net profit attributable to shareholders of 135 million yuan, up 14.7% [1] - The growth in the intelligent computing (智算) business is significant, and its revenue share is expected to continue increasing in the future [1][2] - The gross margin has decreased by 8.25 percentage points to 33.6%, primarily due to the lower margin of the intelligent computing business, but the impact is expected to weaken over time [2] Financial Performance - For Q2 alone, the company's revenue, net profit, and net profit excluding non-recurring items grew by 20.8%, 80.8%, and -31.9% respectively [1] - The revenue breakdown for H1 2025 shows collaborative office products at 250 million yuan (+2.0%), smart channel products at 353 million yuan (+14.8%), and intelligent computing and data intelligence products at 284 million yuan (+50.1%) [1] Business Segments - The collaborative office business is stable, with future growth dependent on the promotion of new technologies and products [1] - The smart channel product line has seen rapid growth, driven by new products such as 5G new calling and AI marketing assistants [1] - The intelligent computing and data intelligence product line has shown remarkable growth due to increased investment in technology services and computing power services [1][2] Cost and Margin Analysis - The gross margin decline is attributed to a 2.4 percentage point drop in the gross margin of technology services and a significant increase in the share of low-margin intelligent computing business [2] - The overall expense ratio decreased from 25.1% in H1 2024 to 22.2% in H1 2025, with sales and management expenses down by 24.0% and 0.5% respectively [2] AI Integration and Future Outlook - The company has developed a comprehensive AI service system, integrating AI infrastructure, model training, and application development [2] - The acceleration of AI penetration in the B-end market is expected to enhance the company's full-stack AI capabilities, driving further growth [2] Profit Forecast - Adjusted profit forecasts for the company indicate net profits of 252 million yuan, 282 million yuan, and 360 million yuan for 2025-2027 [3] - The reasonable valuation level is set at 56 times PE for 2025, corresponding to a target price of 31.36 yuan, maintaining a buy rating [3]

Investment Rating - The report maintains a "Buy" rating for the company, with a target price of 31.36 CNY based on a reasonable valuation level of 56 times PE for 2025 [3][10]. Core Views - The company reported a revenue of 920 million CNY for H1 2025, representing a year-on-year increase of 17.7%, and a net profit attributable to the parent company of 135 million CNY, up 14.7% [2]. - The growth in revenue and net profit for Q2 2025 was particularly strong, with increases of 20.8% and 80.8% respectively, although the non-recurring net profit saw a decline of 31.9% [2]. - The company's AI full-stack system is accelerating, with significant growth in its intelligent computing services and data intelligence product lines, which saw a revenue increase of 50.1% [9]. Financial Summary - The company's projected revenues for 2025-2027 are 2.062 billion CNY, 2.361 billion CNY, and 2.688 billion CNY, with corresponding net profits of 252 million CNY, 282 million CNY, and 360 million CNY [5][10]. - The gross margin is expected to decline to 33.5% in 2025, with a gradual recovery anticipated in subsequent years [5]. - The company has shown a significant reduction in sales and management expenses, contributing to an overall decrease in the expense ratio from 25.1% in H1 2024 to 22.2% in H1 2025 [9]. Business Segments - The revenue from the collaborative office products was 250 million CNY (+2.0%), while the smart channel product line generated 353 million CNY (+14.8%), and the intelligent computing services and data intelligence product line reached 284 million CNY (+50.1%) [9]. - The company is focusing on integrating AI into its product lines, which is expected to enhance its growth potential and drive further revenue increases [9].

Investment Rating - The investment rating for the company is "Buy-A" with a target price of 29.63 CNY, indicating a potential upside from the current price of 26.59 CNY [5]. Core Insights - The company reported a revenue of 920 million CNY for the first half of 2025, representing a year-on-year growth of 17.72%, and a net profit of 135 million CNY, up 14.74% year-on-year. AI and computing-related revenue accounted for 177 million CNY, nearly 20% of total revenue [1][2]. - The company has made a strategic investment in Zhongke Kuyuan Technology, aiming to accelerate the development of atomic quantum computers and promote the industrial application of quantum computing technology [1][3]. Summary by Sections Financial Performance - For the first half of 2025, the company achieved a revenue of 920 million CNY, a 17.72% increase year-on-year, and a net profit of 135 million CNY, reflecting a 14.74% growth [1]. - The company’s AI-related revenue is divided into two main parts: over 100 million CNY from computing-related income and approximately 50-60 million CNY from AI application-related income [2]. Strategic Initiatives - The company has partnered with several funds to invest in Zhongke Kuyuan, which focuses on the research and development of atomic quantum computing technology, aiming to transition from laboratory research to industrial applications [3]. - The company maintains a low debt ratio of 15% and strong cash flow, which supports its strategic investments and business development in AI and computing [3]. Market Position and Future Outlook - The company is positioned to benefit from the growing demand for AI and intelligent solutions across various industries, including partnerships with major clients like China Mobile and China Southern Power Grid [2]. - Revenue projections for 2025-2027 are estimated at 1.98 billion CNY, 2.36 billion CNY, and 2.62 billion CNY, with net profits expected to be 297 million CNY, 366 million CNY, and 402 million CNY respectively [11].

Core Insights - The article emphasizes that Qifu Technology is becoming a key partner in the digital transformation of Chinese banks by integrating financial intelligence with business scenarios [1][7] - Qifu Technology's financial intelligence platform, Deepbank, and its various AI applications are designed to meet the real business needs of banks [2][7] Group 1: Financial Intelligence Development - Qifu Technology launched its self-developed financial intelligence platform Deepbank, which includes four core applications: AI Marketing Assistant, AI Approval Officer, AI Decision Assistant, and AI Compliance Assistant [2][3] - The financial intelligence system is built on a heterogeneous large model platform and a multi-agent collaborative framework, enabling deep understanding of financial semantics and user behavior [2][4] Group 2: Comprehensive Credit Solutions - The upgraded "Qifu Credit Super Intelligent Agent" includes five modules that cover the entire credit business process, enhancing decision-making capabilities [3][6] - The "End-to-End Credit Decision" module utilizes over 700 models and more than 1 billion historical decision data points to accurately assess user risk [3][4] Group 3: Technological Empowerment - Qifu Technology employs four technological engines: data flywheel, multi-modal fusion, self-evolution, and multi-agent collaboration to enhance the intelligence of its financial agents [4][5] - The AI Approval Officer can automatically extract key information from user applications and provide instant recommendations, significantly improving efficiency [5][6] Group 4: Practical Applications and Collaborations - Qifu Technology has demonstrated the effectiveness of its financial intelligence through partnerships with banks, leading to improved customer acquisition and reduced compliance risks [6][7] - A strategic cooperation agreement was signed with Guangdong Huaxing Bank to explore deep applications of AI in credit business, showcasing a collaborative model for digital transformation in banking [6][7]