来源丨凤凰网财经《公司研究院》 作者丨DW （图源：官网） 零一万物是创新工场董事长李开复于2023年创立的AI2.0公司，聚焦大模型技术研发，提供企业级 AI解决方案。 "一把手工程"则是零一万物今年3月"All in toB"战略的核心。该战略强调，企业AI转型必须由CEO 亲自参与战略设计，将AI深度嵌入核心业务流程，才能实现价值交付。这正是CEO李开复所坚信的 路径："唯有CEO亲自参与，依托自研的万智平台，才能实现'以AI交付价值'。" 公开资料显示，沈鹏飞拥有超过26年的IT及互联网行业从业经验，在百度ToB与ToG领域深耕8年， 他曾担任百度智能云中国区副总经理、百度安全副总经理、物联网业务运营总经理等多项核心业务 负责人。 赵斌强出任零一万物AI模型与专业用户产品副总裁，全面负责公司的大模型核心算法研发。他毕业 于清华大学，在互联网算法和AI领域从业17年，曾任阿里巴巴资深算法专家、淘系技术部算法负责 人。 宁宁将担任零一万物国际业务与AI咨询副总裁，全面负责国际市场拓展及企业级、国家级AI转型的 咨询与落地。他博士毕业于新加坡国立大学工程学院，拥有超过二十年的AI与企业服务经验。 10月27 ...

Core Insights - Zero One Technology, one of the "Six Little Dragons" in the domestic large model field, is embarking on a new journey by upgrading its service strategy for government and enterprise sectors, aiming to build a collaborative AI 2.0 ecosystem with partners and industry leaders [2] - The company has faced significant challenges, including high executive turnover, which has raised concerns about its strategic direction and ability to compete in the market [3][4] - The shift in focus from consumer-oriented AI applications to enterprise-level solutions is seen as a necessary survival strategy, although it presents its own set of challenges [4][10] Group 1: Strategic Developments - Zero One Technology announced a comprehensive upgrade of its service strategy at the "Yuanqi Shanghai" conference, leveraging the WanZhi 2.0 platform to create a collaborative ecosystem [2] - The company aims to become an ecosystem connector in the AI era, relying on industry clients for scenarios and data, while partners provide technical capabilities [2] - The strategic shift towards B-end solutions is a response to the competitive landscape and aims to deepen cooperation with leading clients in various sectors [4] Group 2: Executive Changes and Challenges - The company has experienced significant executive turnover, with seven executives, including key technical leaders, leaving the organization [3] - This turnover is attributed to the company's struggles in a competitive market and a strategic retreat to redefine its focus [3][4] - The loss of core technical talent poses risks to the company's research and development capabilities, necessitating a rapid rebuilding of a stable and efficient team [4] Group 3: Market Landscape and Competition - The AI Agent market is projected to grow significantly, with estimates of $5.1 billion in 2024 and $47.1 billion by 2030, indicating a lucrative opportunity for enterprise solutions [6] - However, the competitive landscape is intensifying, with major players like OpenAI, Google, and domestic giants such as Alibaba and Tencent aggressively pursuing the B-end market [7][8] - Zero One Technology's lack of an established ecosystem and scale presents a significant barrier to competing against these larger firms, which have substantial resources and market presence [7][8] Group 4: Future Prospects and Risks - The company is focusing on vertical industries and customized solutions to differentiate itself from larger competitors [8] - However, the path forward is fraught with challenges, including the need to prove long-term value to enterprise clients who may prefer established players [8][10] - The reliance on a few key clients for revenue poses a risk, as larger competitors may target these clients aggressively [10][11]

想要提供私有化部署和定制化的服务，团队的选择是"从一把手抓起"。零一万物联合创始人马杰强 调，在企业AI转型的过程当中，面临的并非简单的技术问题，还有业务融合等管理层面的难题。零一 万物的相关负责人告诉第一财经记者，企业在选择客户的时候也有自己的筛选逻辑，会更倾向于直接 和能够进行决策的一把手等核心管理层进行沟通，以咨询的模式切入业务场景，再实施技术落地。 本文字数：1056，阅读时长大约2分钟 作者 | 第一财经 乔心怡 在大厂依靠"标品化"将Agent落地在平台级企业的另一边，大模型公司"六小龙"之一的零一万物将 目光锁定在B端 Agent的私有化部署红利上。在"'元启上海'华东数智大会"上，第一财经记者采访了 零一万物首次对外亮相的联合创始人沈鹏飞等人了解到，零一万物正试图以定制化和差异化架构路径 切入新的市场。 沈鹏飞认为，不同于大厂的"标品化"方案，更多的细分行业头部企业需要深度的私有化部署服务。 "大厂交付的方式就是把标品扔出来，配合一些培训和服务，客户基于这个标品去做调整。"而这种 模式在实际应用中往往会形成"最后一公里"的难题，"所以大厂能够拿下的项目大多是平台级的，深 入到垂直应用场景的案例 ...

"当数据成为企业的核心资产，私有化部署几乎是必需的选项。"沈鹏飞说，大厂在拿到企业数据之后， 往往会将能力泛化，在相同的行业有所应用。"这是所有的头部公司都不愿意看到的情况。" 想要提供私有化部署和定制化的服务，团队的选择是"从一把手抓起"。零一万物联合创始人马杰强调， 在企业AI转型的过程当中，面临的并非简单的技术问题，还有业务融合等管理层面的难题。零一万物 的相关负责人告诉第一财经记者，企业在选择客户的时候也有自己的筛选逻辑，会更倾向于直接和能够 进行决策的一把手等核心管理层进行沟通，以咨询的模式切入业务场景，再实施技术落地。 除了工程落地之外，架构范式也是Agent在B端落地的难题。 "除了工程落地之外，架构范式也是Agent在B端落地的难题。"零一万物企业解决方案技术负责人王猛解 释，在过去，常见的做法是用一种叫"有向无环图（DAG）"的低代码工作流模式，简单来说就是通过 拖拽和配置来拼装流程。"但这种方式能力有限，已经跟不上大模型时代对灵活、动态智能的需求。" 在大厂依靠"标品化"将Agent落地在平台级企业的另一边，大模型公司"六小龙"之一的零一万物将目光 锁定在B端 Agent的私有化部署红利上 ...

Core Viewpoint - The 2025 Sixth China Fund of Funds Summit highlighted the rapid advancements in AI, particularly the transition from traditional models to intelligent agents, which are expected to significantly enhance business efficiency and create new value in various industries [2][3][4]. Group 1: AI Development Trends - The development of large models has evolved from relying solely on data and computing power to incorporating "slow thinking" capabilities, allowing for deeper reasoning and self-training [3][4]. - The significance of Chinese models, such as Alibaba's Tongyi Qianwen and DeepSeek, lies in their open-source nature, which facilitates easier training and innovation compared to the closed-source models prevalent in the U.S. [3][4]. Group 2: Importance of Intelligent Agents - Intelligent agents are identified as the core future form of AI, possessing memory and execution capabilities that enable them to understand and fulfill business needs, thus acting as "super employees" [4][5]. - The advancement from workflow intelligent agents to reasoning intelligent agents allows for the autonomous breakdown and execution of complex tasks, potentially replacing human labor over hours to days [4][5]. Group 3: Challenges and Strategic Implementation - Traditional enterprises face challenges in deploying intelligent agents, often only replicating past AI capabilities without a clear future strategy [5]. - Successful deployment requires top-level management involvement to align with strategic goals, leading to business restructuring and value redefinition [5]. Group 4: Practical Applications and Value Creation - The company has implemented practical strategies by recruiting experienced consultants to help businesses develop transformation strategies and create quantifiable commercial value through intelligent agents [5]. - Applications in various sectors, such as energy, patent writing, game optimization, and supply chain management, have demonstrated both cost reduction and revenue enhancement [5].