引言：一场被AI催熟的行业洗牌 2025年，中国SaaS行业站在了历史性十字路口。 一边是资本退潮后对盈利模型的严苛审视，一边是生成式AI引爆的技术狂潮。当"AI+"成为所有厂商 PPT首页的标配，市场却已用脚投票：客户不再为"能聊天"的花哨功能买单，而是直指核心——"能省 多少钱？""能带来多少增量价值？" 据IDC最新数据，2025年中国企业级SaaS市场规模达1860亿元，同比增长22.3%，但增速较2024年的 29.7%明显放缓。更值得警惕的是，即便AI SaaS赛道被高调宣称"全面爆发"，其实际ARR（年度经常性 收入）占比仍不足整体市场的15%。大量所谓"AI功能"仍停留在演示Demo阶段，未能转化为真实业务 价值。 "2025年是SaaS行业的'AI压力测试年'。"红杉中国合伙人郑庆生在接受本报独家采访时直言，"很多公 司把AI当作遮羞布，掩盖产品同质化和增长乏力的核心问题。真正的赢家，是那些把AI深度嵌入业务 流、形成不可复制数据飞轮的企业。" 本文聚焦北森、用友、金蝶、泛微、致远等五家代表性上市公司（注："积水潭"并非SaaS企业，疑为误 指；结合上下文或意指医疗信息化企业如卫宁健康，但因 ...

Investment Rating - The report does not explicitly provide an investment rating for the industry. Core Insights - The report emphasizes the rise of AI-native infrastructure as a critical foundation for the digital transformation of industries, driven by the integration of AI technologies across various sectors. This infrastructure is essential for supporting AI-native applications and is seen as a key driver of economic and social transformation in China [8][9]. Summary by Sections 1. Background of AI-Native Infrastructure - The report outlines the increasing demand for AI-native infrastructure due to the ongoing digital transformation and the government's supportive policies. The AI industry in China is projected to exceed 900 billion yuan by 2024, with a year-on-year growth of 24% [23][24]. - The report highlights the shift in AI's role from a mere efficiency tool to a foundational infrastructure akin to electricity, reshaping industry dynamics [19][20]. 2. Development Path and Architecture of AI-Native Infrastructure - The concept of AI-native infrastructure has evolved through three stages: the emergence phase (1950-2009), the exploration phase (2010-2022), and the development phase (2023-present) [31][34]. - AI-native infrastructure is defined as a system designed from the outset to support AI applications, integrating hardware, software, and data to provide comprehensive lifecycle support for AI applications [35][36]. 3. Construction Ideas for AI-Native Infrastructure - The report details various components of AI-native infrastructure, including: - **Intelligent Computing Resources**: A combination of GPU, NPU, and traditional computing resources to support AI model training and inference [45]. - **Unified Scheduling Engine**: A system for dynamic allocation of computing, network, and storage resources tailored to different application scenarios [46]. - **Sandbox**: A secure environment for AI agents to interact with external tools while ensuring system stability [51]. - **Model Development and Production**: A comprehensive toolchain for model tuning, deployment, evaluation, and management [58]. - **Data Supply**: A robust data infrastructure that encompasses data collection, storage, processing, and quality assessment [60][61]. 4. Industry Practice Cases - The report includes various case studies across different sectors, such as telecommunications, government, manufacturing, finance, energy, transportation, and healthcare, showcasing the practical applications of AI-native infrastructure [12][12][12]. 5. Conclusion and Outlook - The report concludes that AI-native infrastructure is poised to become a cornerstone of future economic development, enabling new business models and operational efficiencies across industries [36][37].

Core Viewpoint - The Chongqing Municipal Government has issued an "Artificial Intelligence+" action plan to foster new products and business models in smart consumption [1] Group 1: AI Development Initiatives - The plan encourages the exploration of new business models utilizing AI, aiming to develop a system of smart native technologies, products, and services [1] - There is a focus on developing AI smartphones and computers with multimodal interaction and intent understanding capabilities [1] Group 2: Smart Home Appliances - The initiative promotes the development of AI home appliances such as smart refrigerators and smart laundry machines, aiming to create a comprehensive smart home appliance ecosystem [1] - Solutions involving embodied intelligent robots for scenarios like delivery and shopping are to be introduced, innovating product forms and service models [1] Group 3: Wearable Technology - The plan includes the development of smart wearable products such as AI glasses, watches, and AR/VR devices [1]

Core Insights - OceanBase has surpassed 4,000 global customers since its commercialization in 2020, achieving an average annual growth rate of over 100% for five consecutive years [1] - The company's technology is utilized across more than ten sectors, including finance, government, telecommunications, retail, manufacturing, and the internet, covering 16 countries and regions, over 60 areas, and more than 240 availability zones [1] - The growth is driven by a dual strategy of "private cloud + public cloud," with private cloud customers increasing by 50% in the past year, primarily from core systems with stringent stability and consistency requirements [1][2] - OceanBase has maintained the top position in the Chinese market for local deployment distributed databases for three consecutive years and is the leader in the local deployment market for distributed finance in China [1] Public Cloud Growth - OB Cloud has seen an average annual customer growth of 115% over the past three years, with revenue contribution reaching 30% [1][2] - The core advantage of OB Cloud lies in its "multi-cloud native" capabilities, enabling OLTP and storage separation based on object storage, ensuring business continuity through cross-availability zone disaster recovery and automatic switching [2] International Expansion - OceanBase's vector database capabilities have entered the global top ten in DB-Engines, with 16 papers accepted at major database conferences in 2025 [2] - In Southeast Asia, OB Cloud has served local fintech companies such as Indonesia's DANA and Singapore's Atome, enhancing peak load capacity by ten times for GoTo's risk control engine [2] Data and AI Integration - OceanBase has launched a new paradigm of "Data×AI," integrating transaction processing, analytical processing, and AI capabilities into a single core with the release of version 4.4 [3] - The company has open-sourced its first AI-native hybrid search database, seekdb, which supports unified retrieval of vector, full-text, scalar, and GIS data, compatible with over 30 mainstream AI frameworks [3][4] - The strategic focus is on creating an integrated data foundation for the future, emphasizing the fusion of architecture, storage, and load, while increasing global open-source investment [3][4] Future Directions - The company aims to support both humans and intelligent agents within a single database, driving real-time and trustworthy intelligence from the data source [4] - The open-sourcing of seekdb and the launch of the new domain oceanbase.ai are key steps in promoting the industry's transition to an "intelligent native" architecture [4]

Core Insights - The article highlights the paradox of high consumption and low returns in the AI industry, emphasizing that 95% of generative AI investment projects fail to deliver expected financial returns, with only 5% achieving commercialization [1][4] - Beijing Zhongguancun KJ Technology Co., Ltd. is positioned as a leading player in the enterprise-level AI model application market, having established a strong foothold by focusing on vertical applications rather than chasing trends [1][3][4] Market Dynamics - By mid-2025, the daily consumption of enterprise-level AI models in China is projected to reach 10.2 trillion tokens, equivalent to 46 billion 2,000-word articles, indicating a massive demand for AI solutions [1] - The article discusses the shift from a "technology showcase" era to a focus on "value realization" in AI, where deep engagement in vertical sectors is essential for successful AI integration [1][4] Company Strategy - Zhongguancun KJ's strategy began with a "reverse layout" in 2014, focusing on intelligent audio and video technology instead of mainstream computer vision, which has become a core asset for connecting businesses with customers [4] - The company has strategically chosen to concentrate on enterprise-level intelligent interaction scenarios, particularly in the smart customer service sector, which is seen as a critical entry point for large model applications [4][12] Competitive Position - In the latest IDC report, Zhongguancun KJ ranks fourth in the Chinese intelligent customer service market, leading among AI model companies [5] - The company’s approach emphasizes that the winners in the AI arms race will be those who can translate model capabilities into commercial value, rather than merely possessing the largest models [6] Implementation Framework - Zhongguancun KJ has proposed a "platform + application + service" three-tier engine strategy to accelerate the deployment of vertical AI models, addressing core issues of usability and effectiveness in enterprise applications [13][16] - The company aims to create a closed-loop system that activates enterprise data assets, integrates various AI capabilities, and continuously optimizes performance through iterative feedback [12][16] Industry Applications - The article provides examples of successful collaborations across various sectors, including finance, manufacturing, and infrastructure, showcasing how Zhongguancun KJ's AI models enhance operational efficiency and knowledge transfer [18][19][21][22] - Notable projects include a training platform for securities firms that improves training efficiency by 70% and a model for the shipbuilding industry that enhances intelligence analysis efficiency by 60% [19][21] Conclusion - The article concludes that the true value of AI lies not in the amount of computational power used but in the ability to understand and address industry-specific challenges, marking a shift from theoretical to practical applications in AI [25][26]

Core Viewpoint - AI represents the last opportunity for small and medium-sized enterprises (SMEs) to enhance their business efficiency and cash flow, especially as larger companies leverage AI to eliminate their operational disadvantages [4][5][52]. Group 1: AI's Impact on SMEs - SMEs have a lower organizational complexity, shorter decision-making chains, and lighter IT burdens, making them more agile in transforming AI capabilities into business efficiency [4]. - The introduction of AI allows SMEs to potentially outpace larger companies in specific niches before those companies can adapt [29][52]. - The current technological wave favors SMEs as the barriers to entry have shifted from technology to organizational adaptability [10][20]. Group 2: Strategic Recommendations for SMEs - SMEs should focus on restructuring their operations to make AI the default executor, with human roles limited to decision-making and exceptions [16][50]. - Prioritizing end-to-end automation is crucial, moving beyond isolated applications to fully integrated processes [16]. - SMEs should aim for deep specialization in niche markets, leveraging their understanding of data and processes rather than solely relying on the most powerful models [16]. Group 3: Competitive Landscape - Once large enterprises fully integrate AI into their processes, their scale advantages will extend into previously inaccessible areas, leading to increased market concentration [25][26]. - The risk for SMEs lies in their reliance on low-value-added processes, which AI and automation can easily disrupt [27][28]. - SMEs must either establish a stronghold in niche markets or risk being outpaced by larger firms that have streamlined their operations through AI [29][52]. Group 4: Implementation Roadmap - A phased approach is recommended for SMEs, starting with pilot projects and gradually moving towards full automation [42][44]. - Key performance indicators (KPIs) should be established to measure automation coverage, service costs, and resolution rates to ensure continuous improvement [38][39][40]. - The transition to an "intelligent native" organization requires minimizing friction between data and processes, and rethinking business structures to fully leverage AI capabilities [50][51].

Group 1 - The Monterey Institute of International Studies (MIIS), a prestigious translation school, announced it will stop enrolling graduate students by June 2027, marking a significant decline in the translation education sector [1][4][5] - The decline in enrollment is attributed to financial difficulties and a drastic reduction in demand for human translators due to advancements in AI translation technology, which has improved efficiency by nearly nine times and reduced costs by over 90% [5][6] - The impact of AI is not limited to translation; it is reshaping various industries, with around 20% of Chinese universities adjusting their programs in response to AI's influence, indicating a potential decline in the relevance of English as a major [8][9] Group 2 - The Chinese government has launched a comprehensive AI strategy, with a focus on integrating AI into key sectors by 2027, aiming for over 70% penetration of new intelligent devices [12][13] - The AI strategy outlines a three-step timeline, with AI expected to become a major economic driver by 2030 and fully integrated into society by 2035, highlighting the urgency for industries to adapt [12][13] - The concept of "intelligent native" ecosystems is introduced, where products and business models are fundamentally driven by AI, indicating a shift from platform-centric to user-intent-centric approaches [15] Group 3 - The AI era is anticipated to transform urban landscapes, with cities that possess strong research capabilities or application innovation likely to gain competitive advantages [16][17] - Major cities like Beijing and Shanghai are positioned to lead in AI development due to their academic resources, while cities like Shenzhen and Hangzhou can leverage their industrial bases for rapid AI application [17][18] - Emerging cities may also carve out niches in specific AI applications, showcasing the diverse potential for growth in the AI landscape [18][20]

Core Viewpoint - The article emphasizes that the true competition in AI is the "landing efficiency" of applications, highlighting the ongoing "infrastructure battle" regarding AI middleware [2][6]. Group 1: Event Details - A live broadcast is scheduled for September 23, from 20:00 to 21:30, focusing on the transition from "model-centric" to "application-centric" approaches in AI middleware [2]. - The event will feature experts from the industry, including a senior technical expert from Ant Group and the CTO of Memory Tensor [3]. Group 2: Key Challenges - The article raises questions about how enterprises can transition smoothly from "cloud-native" to "intelligent-native" systems [3]. - It discusses the challenges developers face in capturing the current opportunities and becoming core talents in the intelligent era [6]. Group 3: Live Broadcast Content - The live session will cover topics such as the engineering framework for Agent applications and practical implementations of the RAG framework [7]. - Participants will have the opportunity to ask questions to the instructors during the live session [8].

Core Insights - The article discusses the inefficiencies and structural delays present in both the digital and physical worlds, highlighting the need for a more seamless integration of personal agents and automated production systems [5][10][12]. Digital World Analysis - The current internet is described as a collection of isolated data islands rather than a truly interconnected network, leading to significant friction and inefficiencies in data access and usage [5][7]. - The concept of a personal agent is introduced as a necessary tool for individuals to navigate the digital landscape, but existing tech giants create barriers that prevent optimal data flow and user experience [6][8]. - The article posits that the existing structure based on data monopolies is outdated and will inevitably collapse under the pressure of personal agents seeking to optimize user experiences [7][8]. Physical World Analysis - In the physical realm, the article identifies similar structural inefficiencies, where supply chains operate as isolated entities, leading to delays and resource wastage [10][11]. - The current manufacturing processes are criticized for being based on incomplete data, resulting in poor market predictions and unnecessary production [11]. - A vision for a "smart-native" physical world is presented, where user demands can be instantly transformed into production instructions, eliminating the need for traditional inventory and logistics [12][14]. Future Vision - The concept of "无人公司" (unmanned companies) is introduced, which are AI-driven entities that operate without traditional management structures, responding directly to user needs through automated processes [13][14]. - The article envisions a future where the connection between human intent and physical production is instantaneous, facilitated by personal agents and unmanned companies, thus eliminating inefficiencies in the current system [17]. - The transformation is framed as a shift from a world of translation and delays to one where desires can be directly realized, marking a significant evolution in how humans interact with technology and production [16][17].

Group 1 - The core idea is that AGI (Artificial General Intelligence) is not a future concept but is already present and evolving in the current environment [1][11][64] - The emergence of "intelligent native" companies is highlighted, which signifies a shift in how technology and organizational models interact [5][8][12] - The concept of "intelligent native" is described as a value creation system where AI becomes the primary agent, simplifying traditional organizational processes [29][30] Group 2 - The rapid evolution of AI is emphasized, with current AI capabilities being significantly advanced compared to those in 2022 [17][18] - The traditional software development process is contrasted with the "intelligent native" approach, which streamlines collaboration and enhances productivity [24][25][27] - The recursive nature of organizational and business structures is discussed, indicating that as AI capabilities grow, the complexity of organizations can be reduced [31][39] Group 3 - The need for a new paradigm in value creation is stressed, as AI technology becomes more accessible and its application more critical [44][46] - The concept of "无人公司" (Unmanned Company) is introduced, suggesting a future where companies operate with minimal human intervention, driven by AI [50][62] - The importance of redefining roles and processes in light of AI advancements is highlighted, indicating that success will depend on adapting to these changes [64][65]