广告
加载中

AI应用的最后一公里 产业公司接棒

张睿 2026-09-11 17:24
张睿 2026/09/11 17:24

邦小白快读

EN
全文速览

当前AI行业已出现核心转向,以聊天为核心的Chatbot模式告一段落,能实际办事的智能体时代正式来临,普通人可把握以下关键信息与实用认知。

1. 应用进展:国内WorkBuddy、千问办公、豆包工作等AI办公产品已落地职场,可协助处理文档、代码开发、创意设计等日常工作,目前类似“实习生”角色,待补足行业场景、数据、流程经验后将成为能独当一面的“行业老兵”。

2. 实用认知:企业无需走先信息化、再数字化、再智能化的传统升级路径,拥有大量业务对话数据的企业可借助成熟智能体产品直接跃迁到AI原生阶段;普通人使用AI不能停留在聊天提问层面,要选择能嵌入真实工作流程、对应岗位功能的产品,才能真正提升效率。

3. 行业常识:2026年国内企业级AI Agent市场规模将突破480亿元,同比增速接近300%,未来用好AI做数据驱动的精益工作是必选项,不是可选项。

AI产业已进入落地交付的关键阶段,企业级智能体成为商业化核心赛道,品牌布局To B类AI业务可重点参考以下方向。

1. 产品研发逻辑:AI产品不能停留在通用聊天问答层面,要围绕垂直场景补足行业知识、业务流程、经验数据短板,不仅要学习企业ERP等系统里的结果数据,还要归集微信、电话、邮件等渠道的沟通过程数据,让AI真正能理解任务、调用工具、执行流程,打造对应采购、销售、运营等真实岗位的“AI同事”类功能。

2. 合作与渠道思路:无需从零研发底层大模型等通用AI能力,可与成熟通用AI平台合作,将自身行业经验叠加到通用智能体底座上,降低研发成本;可推出生态合作伙伴计划,通过渠道合作、产业平台合作等模式降低客户接入门槛,扩大覆盖范围。

3. 定价参考:可采用订阅费+Token消耗费+交易服务费的混合定价模式,从售卖工具、收取订阅费转向交付业务结果、按创造的价值收费。

智能体时代B2B贸易领域迎来越级发展的新窗口,贸易类卖家可把握政策与市场红利,找准增长路径,规避发展误区。

1. 机会提示:工信部正开展人工智能应用服务商培育专项行动,2026年国内企业级AI Agent市场规模将突破480亿元,同比增速近300%,贸易流通行业有数十万亿市场规模但AI化程度极低,属于蓝海赛道;2026年上半年企业Agent间的A2A交易已带动GMV11.05亿元,同比增长73.2%,未来“Agent找Agent”将成为B2B交易新范式。

2. 落地路径:无需重复走完信息化、数字化、智能化的长周期升级路径,作为对话驱动型的贸易商家,可借助成熟的垂直领域AI Agent产品直接跃迁到AI原生阶段,不用从零搭建技术体系。

3. 风险提示:要避开AI使用误区,不能仅把聊天问答当作用上AI,要接入能覆盖寻源、比价、报价、合同、物流、结算全流程的AI工具;当前B2B粗放式增长已彻底结束,用好AI实现数据驱动的精益管理是必选项,不是可选项。

智能体时代工厂可依托产业AI工具大幅降低数字化转型成本,在供应链管理、供需对接等环节挖掘效率提升空间,把握新的商业机会。

1. 数字化转型新路径:无需按部就班走完信息化、数字化再到智能化的传统升级路线,可借助成熟的垂直类AI Agent产品,直接接入采购、库存、运营、销售等环节的AI能力,轻装上阵完成智能化升级,减少不必要的系统建设投入。

2. 生产经营效率提升方向:可使用匹配贸易流通场景的AI岗位工具,解决原材料采购寻源、供应商比价、订单库存管理、行情监测、经营报告生成等环节依赖老员工经验、流程不标准、数据分散在微信/电话/邮件等渠道的问题,逐步将散落的多源业务数据按AI可学习的逻辑存储,支撑AI智能处理业务。

3. 合作机会:可对接NextB2B这类产业AI生态,参与到A2A交易体系中拓宽供需对接渠道,该类AI能力跑通后可复制到化工、汽配、货代等多类工业流通领域,工厂可提前适配相关规则。

AI商业化已从技术竞赛转向落地交付阶段,企业级智能体成为核心增长赛道,AI服务商可结合自身定位找准客户痛点,打造适配的解决方案。

1. 行业趋势参考:当前AI产业投资逻辑已从模型能力比拼、资本开支竞赛,转向订单、收入、盈利的商业化验证,Chatbot仅占AI整体市场的十分之一,2026年国内企业级AI Agent市场规模将突破480亿元,同比增速近300%;资本正向有真实场景和数据的垂直玩家倾斜,缺少稳定商业化能力的通用Agent初创项目正遭遇融资遇冷的情况。

2. 客户核心痛点:大量实体、贸易类企业存在明显AI落地壁垒,一是通用AI不了解行业规则、业务流程与客户真实需求,只会聊天不会实际干活;二是多数中小企业数字化基础薄弱,员工数字化能力参差不齐,难以承担从零搭建AI体系的成本,复杂AI工具落地难度大;三是行业过程数据分散在微信、电话、邮件等多个渠道,未按AI可学习的逻辑归集,无法支撑AI执行具体任务。

3. 解决方案方向:可走通用技术+产业能力合作的路线,通用技术厂商做好Agent基础设施,产业服务商沉淀行业技能、工作流、数据经验,打造开箱即用的岗位专属AI产品,通过生态合作降低客户接入门槛。

AI落地最后一公里阶段,产业互联网平台的核心定位正从交易撮合者转向产业AI应用平台,平台可调整业务布局、运营模式与合作体系,抓住新的增长曲线。

1. 风向与需求变化:B2B领域粗放式增长已彻底结束,平台商家普遍存在精益化管理、降本提效的强需求,需要能直接嵌入业务流程、开箱即用的AI工具,不愿承担从零搭建数字化体系的高成本;工信部当前正重点培育懂行业痛点、通技术机理、知安全风险、善交付运营的AI应用服务商,平台过去积累的场景、数据、流程经验正从成本中心转为核心资产。

2. 可参考的落地实践:找钢网与腾讯WorkBuddy合作推出NextB2B产品,没有盲目投入底层大模型研发,而是将自身十余年积累的贸易场景经验、数据、工作流转化为AI能力,设置采购、销售、内勤、运营等岗位专属AI模块,帮用户完成全流程业务任务。

3. 运营与招商方向:可推出生态合作伙伴计划,通过渠道合作、跨产业平台合作扩大覆盖,降低商家AI接入门槛;商业模式可从传统的订阅费、交易佣金,转向订阅费+Token消耗费+A2A交易服务费的混合模式,提前布局Agent间交易的运营规则,抢占新交易范式的先机。

2026年AI产业进入智能体落地的关键转折期,产业互联网公司成为打通AI落地最后一公里的核心主体,产生了诸多值得研究的产业新动向、路径创新与模式变革。

1. 产业新动向:AI发展核心路径从对话式Chatbot转向能执行任务的智能体,投资逻辑从模型能力竞赛、资本开支比拼转向商业化落地的订单、收入与盈利验证;2026年国内企业级AI Agent市场规模将突破480亿元,同比增速近300%,资本正向拥有真实场景与数据的垂直玩家集中,无稳定变现能力的通用Agent项目融资遇冷;产业互联网平台定位从交易撮合转向产业AI应用平台,过往积累的行业经验、流程、数据成为核心资产。

2. 新路径与新问题:中国AI落地或将走出不同于美国的路径,由通用入口平台与垂直产业应用公司共同完成最后一公里交付;B2B企业无需遵循信息化-数字化-智能化的线性升级路径,对话驱动型企业可依托智能体直接越级跃迁到AI原生阶段;当前落地仍面临通用模型缺乏产业经验、产品易用性不足、数据安全待保障、商业模式待验证等问题。

3. 商业模式创新:AI服务正从MaaS向AaaS演进,收费模式从卖工具转向按交付价值收费,出现订阅费+Token消耗费+A2A交易服务费的混合模式,“Agent找Agent”的范式有望重构B2B交易逻辑。

返回默认

声明:快读内容全程由AI生成,请注意甄别信息。如您发现问题,请发送邮件至 run@ebrun.com 。

我是 品牌商 卖家 工厂 服务商 平台商 研究者 帮我再读一遍。

Quick Summary

The AI industry is undergoing a core strategic shift: the chat-centric chatbot era is drawing to a close, and the age of task-capable AI agents has officially arrived. Key takeaways for the general public are as follows:

1. Application progress: Domestic AI workplace products including WorkBuddy, Qianwen Office, and Doubao Work have already been deployed across professional settings, supporting daily tasks such as document processing, software development, and creative design. These tools currently operate at the capability level of an intern; once they accumulate sufficient industry-specific scenario, data, and workflow expertise, they will evolve into reliable, high-performing "veteran contributors" capable of independent end-to-end work.

2. Practical guidance: Enterprises do not need to follow the traditional linear upgrade path of IT building, digitization, and then intelligentization. Companies with large volumes of business conversation data can leverage mature agent products to leap directly to the AI-native stage. For individual users, AI engagement cannot stop at chat-based queries; to drive real efficiency gains, users need to select products that integrate directly into real work workflows and map to specific job functions.

3. Industry baseline: China’s enterprise AI agent market is projected to exceed RMB 48 billion by 2026, with a year-over-year growth rate of nearly 300%. Leveraging AI to enable data-driven, lean work practices will become a necessity, not an optional choice, going forward.

The AI industry has entered the critical stage of real-world deployment and delivery, with enterprise-grade agents emerging as the core commercialization track. Brands developing B2B AI offerings can reference the following strategic directions:

1. Product R&D logic: AI products cannot remain at the stage of generic chat-based Q&A. They need to fill gaps in vertical industry knowledge, business workflows, and empirical data: instead of only learning outcome data from internal systems such as ERP, products must also collect unstructured process data from communication channels including WeChat, calls, and emails, to enable AI to truly understand tasks, call relevant tools, and execute end-to-end workflows. This will power "AI colleague" functions tailored to real roles such as procurement, sales, and operations.

2. Partnership and channel strategy: Brands do not need to build general-purpose foundational large language models from scratch. Instead, they can partner with mature general AI platforms, overlay their own industry expertise onto universal agent infrastructure, and reduce R&D costs. They can also launch ecosystem partner programs, lowering customer onboarding barriers and expanding market reach via channel collaboration and industry platform partnerships.

3. Pricing reference: Brands can adopt a hybrid pricing model combining subscription fees, token usage charges, and transaction service fees, shifting from selling tools and charging recurring subscriptions to delivering business outcomes and pricing based on the value created for customers.

The AI agent era is opening a new window for leapfrog development in the B2B trade sector. Trade sellers can capture policy and market dividends by identifying clear growth paths and avoiding common development pitfalls.

1. Opportunity outlook: China’s Ministry of Industry and Information Technology (MIIT) has launched a special initiative to cultivate high-quality AI application service providers. The domestic enterprise AI agent market is set to exceed RMB 48 billion by 2026, growing nearly 300% year-over-year. The trade and circulation sector, which totals tens of trillions of RMB in market size but has extremely low AI penetration, represents a clear blue-ocean track. In the first half of 2026, agent-to-agent (A2A) transactions between enterprise agents already drove RMB 1.105 billion in GMV, up 73.2% year-over-year; "agent sourcing agent" will become a new paradigm for B2B trade going forward.

2. Implementation path: Trade sellers, as conversation-driven businesses, do not need to go through the long, linear upgrade cycle of IT building, digitization, and then intelligentization. They can leverage mature vertical-specific AI agent products to leap directly to the AI-native stage, without building an in-house technical system from the ground up.

3. Risk reminder: Sellers must avoid common AI usage pitfalls: chat-based Q&A alone does not count as meaningful AI adoption. Teams need to adopt AI tools that cover the full trade workflow, including sourcing, price comparison, quotation, contract management, logistics, and settlement. The era of extensive, high-growth B2B operations is fully over; leveraging AI to achieve data-driven lean management is an imperative, not an optional choice.

In the AI agent era, manufacturers can leverage industry-specific AI tools to substantially cut digital transformation costs, unlock efficiency gains across supply chain management and supply-demand matching, and capture new business opportunities.

1. New digital transformation path: Factories do not need to follow the traditional step-by-step upgrade route of IT construction, digitization, and then intelligentization. Instead, they can use mature vertical AI agent products to directly access AI capabilities for procurement, inventory management, operations, and sales, completing intelligent upgrades with light investment and avoiding unnecessary system construction costs.

2. Operational efficiency improvement priorities: Factories can deploy AI role-specific tools tailored to trade and circulation scenarios to address longstanding pain points, including overreliance on veteran employee experience for raw material sourcing, supplier price comparison, order and inventory management, market trend monitoring, and business report generation, as well as unstandardized workflows and scattered business data across WeChat, calls, and emails. Teams can gradually store fragmented multi-source business data in structures that AI can learn from, to support AI-powered autonomous business processing.

3. Partnership opportunities: Factories can connect to industry AI ecosystems such as NextB2B, and participate in the A2A transaction system to expand supply-demand matching channels. Once these AI capabilities are proven, they will be replicable across multiple industrial circulation segments including chemicals, auto parts, and freight forwarding, and factories can adapt to relevant rules in advance.

AI commercialization has shifted from technology competition to real-world deployment and delivery, with enterprise-grade agents becoming the core growth track. AI service providers can align their positioning with customer pain points to build fit-for-purpose solutions.

1. Industry trend reference: Investment logic in the AI sector has shifted from competing on model capabilities and capital expenditure to validating commercial viability via orders, revenue, and profitability. Chatbots represent only one-tenth of the total AI market; China’s enterprise AI agent market is projected to exceed RMB 48 billion by 2026, growing nearly 300% year-over-year. Capital is flowing toward vertical players with real-world scenarios and proprietary data, while general-purpose agent startups without stable commercialization capabilities are facing a funding chill.

2. Core customer pain points: A large number of real-economy and trade enterprises face clear barriers to AI adoption. First, generic AI tools lack understanding of industry rules, business workflows, and real customer needs, and can only chat rather than complete practical work. Second, most SMEs have weak digital foundations and uneven employee digital literacy, making it difficult to bear the cost of building an AI system from scratch, and complex AI tools are hard to roll out. Third, industry process data is scattered across channels such as WeChat, calls, and emails, and is not organized in structures that AI can learn from, making it impossible to support AI execution of specific tasks.

3. Solution direction: Providers can adopt a model of collaboration between general technology and industry capabilities: general technology vendors can build robust agent infrastructure, while industry service providers codify vertical expertise, workflows, and data insights to build out-of-the-box, role-specific AI products, lowering customer onboarding barriers via ecosystem collaboration.

As AI moves into the last mile of real-world deployment, the core positioning of industrial internet platforms is shifting from transaction matchmaker to industrial AI application platform. Platforms can adjust their business layout, operating models, and partnership systems to capture new growth curves.

1. Trend and demand shifts: The era of extensive growth in the B2B sector is fully over. Merchants on platforms have strong demand for lean management, cost reduction, and efficiency improvement, and need out-of-the-box AI tools that integrate directly into business workflows, rather than bearing the high cost of building digital systems from scratch. MIIT is currently prioritizing the cultivation of AI application service providers that understand industry pain points, master technical mechanisms, are aware of security risks, and excel at delivery and operations. The scenario, data, and workflow experience that platforms have accumulated over years of operation is shifting from a cost center to a core asset.

2. Reference implementation case: Zhaogang.com partnered with Tencent WorkBuddy to launch the NextB2B product. Instead of investing blindly in foundational large model R&D, the team converted its more than 10 years of accumulated trade scenario expertise, data, and workflows into AI capabilities, building dedicated AI modules for roles including procurement, sales, back-office support, and operations to help users complete full-cycle business tasks.

3. Operations and merchant recruitment direction: Platforms can launch ecosystem partner programs to expand coverage via channel collaboration and cross-industry platform partnerships, lowering the barrier for merchants to adopt AI. For business models, platforms can shift from traditional subscription fees and transaction commissions to a hybrid model of subscription fees, token usage charges, and A2A transaction service fees, designing operating rules for inter-agent transactions in advance to gain first-mover advantage in the new transaction paradigm.

In 2026, the AI industry is entering a critical inflection point for agent deployment. Industrial internet companies are emerging as the core actors bridging the last mile of AI implementation, giving rise to notable new industry trends, path innovations, and business model transformations worthy of research.

1. Emerging industry trends: The core development path of AI is shifting from conversational chatbots to task-executing agents, while investment logic is moving beyond model capability races and capital expenditure competition to validation of commercial viability via orders, revenue, and profitability. China’s enterprise AI agent market is projected to exceed RMB 48 billion by 2026, growing nearly 300% year-over-year. Capital is concentrating among vertical players with real-world scenarios and proprietary data, while general-purpose agent projects without stable monetization pathways face cooling investor interest. The positioning of industrial internet platforms is evolving from transaction matchmakers to industrial AI application platforms, with their accumulated industry expertise, workflows, and data becoming core assets.

2. New implementation paths and open questions: China’s AI deployment is likely to follow a path distinct from the US market, with general-purpose entry platforms and vertical industry application firms jointly delivering last-mile implementation. B2B enterprises do not need to follow the linear upgrade path of IT building, digitization, and then intelligentization; conversation-driven enterprises can use agents to leap directly to the AI-native stage. Current deployment barriers remain, including foundational models’ lack of industry expertise, insufficient product usability, unresolved data security concerns, and unproven business models.

3. Business model innovation: AI services are evolving from MaaS (Model as a Service) to AaaS (Agent as a Service). Pricing models are shifting from tool sales to value-based delivery, with the emergence of hybrid models combining subscription fees, token usage charges, and A2A transaction service fees. The "agent sourcing agent" paradigm has the potential to restructure core B2B transaction logic.

Disclaimer: The "Quick Summary" content is entirely generated by AI. Please exercise discretion when interpreting the information. For issues or corrections, please email run@ebrun.com .

I am a Brand Seller Factory Service Provider Marketplace Seller Researcher Read it again.

【亿邦原创】“AI的下一站,不是会聊天的模型,而是会行动的智能体。”Google I/O开发者大会上,Google CEO桑达尔·皮查伊(Sundar Pichai)如是直言。

2026年,AI行业正在经历一场深刻的转向:以“聊天”为核心的Chatbot模式告一段落,以“办事”为能力的智能体时代已然来临。

在国内市场,WorkBuddy、千问办公、豆包工作等AI办公产品面世几个月,已经进入不同的工作环境,像一名勤奋的“实习生”一样,协助职场人解决文档处理、代码开发、创意设计等日常工作问题。

那么“实习生”如何变成经验丰富的“行业老兵”?

这正是企业级AI落地的难点:AI能开发代码,但它不了解这家客户的需求;AI能做PPT,但它不知道这个行业的生意策略。换句话说,AI进入真实产业,亟需补足的不是技术,而是场景、数据、流程、经验。

而这些,在产业互联网公司手中。

9月10日,找钢网集团在以“产业AI,落地生根”为主题的第二届贸易企业AI应用论坛上,联合腾讯WorkBuddy正式发布面向B2B贸易场景的AI Agent产品NextB2B,并启动NextB2B生态合作伙伴计划。NextB2B将持续开放产品、技术与产业能力,通过渠道合作、产业互联网平台合作等多种模式,进一步降低企业AI化门槛,让更多产业伙伴不必从零开始,更快进入AI原生阶段,共享产业AI带来的新业务与新增长。

这意味着,以找钢网集团为代表的产业互联网公司,正在将过去十余年积累的数据经验,提炼为AI能力,向上下游、跨行业企业输出。此举也意味着,产业互联网在智能体时代,找到了新的价值。

01

AI走向应用落地,对话驱动型企业迎来越级发展的机会

8月31日,工信部发布《关于开展人工智能应用服务商培育专项行动的通知》,提出培育一批“懂行业痛点、通技术机理、知安全风险、善交付运营”的人工智能应用服务商。

国家层面明确将“应用服务商”定位为AI产业链的关键角色,而产业互联网公司在其中天然具备优势。

产业互联网公司深耕贸易流通行业,拥有大量非标准化、强经验、强流程的业务场景。以钢铁贸易为例,钢材品类规格繁多,每笔交易的品质、价格、交付条件都不同;找货比价依赖采购员的人脉和经验,报价判断依赖销售对市场的直觉;一笔交易涉及寻源、比价、报价、合同、订单、物流、结算等多个环节,每个环节都有行业特有的规则和风险点。

AI如何在这些场景中落地?首先要解决的是数据问题,数据又分为过程数据和结果数据。找钢网创始人、董事长兼CEO王东曾说:“先有信息流,再有贸易流。”信息流就是过程数据——跟客户怎么询盘、怎么议价、怎么促进成交;贸易流是结果数据——存储在ERP、电商系统、财务软件中,用来反映公司的经营成果。

过往的企业数字化建设,主要体现在结果数据上,如果只让AI学习结果数据,那它就只会数据分析,如果想让AI学会干活,那它必须学习过程数据。

贸易流通行业的过程数据是分散的,很多交易发生在微信里——需求单、合同、打款底单、发票、对账单都在微信里传递,还有交易发生在电话、邮件,以及面对面的谈判里。

过去十年,不同行业的产业互联网平台,以交易为纽带,将产品、合同、物流、金融等数据在线化、标准化,已经做了一部分工作,但是还远远不够,接下来真正需要建设的是,把贸易流通场景下散落的多源数据,按AI可学习的逻辑存储起来,积累足够的上下文支撑AI智能处理业务。

找钢网集团管理合伙人兼副总裁张晓坤对此表达了乐观态度:“B2B贸易企业天然就是对话驱动的。很多业务协同发生在微信、IM和各种消息里。过去我们总觉得,这些企业信息化、数字化基础不够重,是一个短板。但到了AI时代,它反而可以轻装上阵。”

在他看来,过去二十多年,很多企业走的是先信息化、再数字化、再智能化的路径,要先建系统、沉淀数据、重构流程,再一步一步走到AI。但今天,Agent天生能够理解对话、理解上下文、调用工具、完成任务。因此拥有大量业务对话上下文的B2B企业,不必把过去二十年的路重新走一遍,可以借助成熟的Agent OS、产业Know-how和NextB2B这样的产品,直接向AI原生阶段跃迁。

“打个比方,过去是从DOS走到Windows,再一步步走到AI;今天,一些B2B企业有机会直接跳到AI。我们把它叫做——越级进化。”张晓坤说。

这意味着,AI时代,千行百业在某种意义上重新站到了同一起跑线上。谁能更快把真实业务、行业经验和AI结合起来,谁就可能率先拿到新的增长机会。

02

产业公司的解法:经验AI化,能力产品化

在真正的实业产业里,大大小小的工厂、贸易企业,似乎跟AI技术之间“有壁”。有些老板以为打开豆包,问几个问题,就相当于“使用AI”了。所以关键的问题在于,AI懂不懂业务,有没有融入企业,能不能真正干活。

“做贸易流通、做B2B的这帮人,学历参差不齐,也不全都是年轻人。你怎么让他们用上最先进的AI?最好的办法就是把实施过的客户经验固化成一个版本,直接让他们开箱即用。”张晓坤说。

找钢网给出的答案是NextB2B——基于腾讯WorkBuddy的B2B全流程AI Agent解决方案。传统聊天机器人以你问我答、信息检索为主,而NextB2B是理解任务、调用工具、执行流程、交付结果。其底层逻辑是:腾讯提供通用AI智能体能力,找钢网提供泛贸易流通行业的SKILL、MCP、数据和工作流,将通用AI能力转化为产业应用能力。

用户可以在WorkBuddy内切换到NextB2B贸易通版本,看到的就不是日常办公、代码开发这些通用功能,而是采购、销售、内勤、运营这些贸易岗位的专属界面。每个模块对应一个真实业务角色:采购专家负责寻源与供应商比价,销售专家负责需求识别与报价生成,内勤专家负责订单库存管理与风险查询,运营专家负责经营报告生成与行情监测。而且,每个模块不是泛泛的聊天助手,而是“AI同事”。

NextB2B不仅能应用于钢贸流通行业,也可以应用于其他品类流通行业。那么,为什么一个深耕钢铁贸易十余年的产业公司,要试图解决“AI如何进入B2B贸易流通”的问题?

“我们自己被虐得足够多,场景我们接得比较多。”张晓坤坦言。找钢网2025年集团交易量超过1500亿元,每天有1900多万次业务交互。找钢网不只是电商,还拥有第三方物流平台、B2B支付、供应链金融等复杂业务场景。而且,找钢网2018年与腾讯合资的公司还服务了6300多家非钢企业、覆盖20多个行业,这些经验被高度抽象,沉淀出通用的B2B解决方案。

实际上,拥有产业经验的公司,将数据经验构建为AI能力向外输出,也有多种形式。有的在原有软件内植入AI数字员工,有的开发了全新的行业AI工作台,但找钢网选择了与腾讯合作,在WorkBuddy的基础上叠加行业经验,相当于将通用能力与产业垂直能力相乘。

张晓坤的考量很直白:“我们应该去做电器,而不是想办法去做发电机,或者去铺国家电网。我们就踏踏实实做出更多的电器出来就行。”也就是说,WorkBuddy封装好了从产品交互层、Agent Harness到底层大模型调用的一整套基础设施。对产业公司而言,重新研发这些通用能力既不擅长,也不经济。产业公司应该把重心投入到行业知识、业务流程、数据经验在智能体时代的转化与应用上去。

03

产业互联网公司在Agent时代的新起点

中信建投证券研报称,AI产业投资逻辑正从模型能力与资本开支竞赛,逐步转向订单、收入及盈利的商业化验证。

昆仑万维周亚辉直言:Chatbot仅占AI整体市场的十分之一,企业级Agent正成为AI商业化的“核心金矿”。2026年国内企业级AI Agent市场规模突破480亿元,同比增速接近300%。

但资本也在分化——大量缺少稳定商业化变现能力的通用型Agent初创项目融资遇冷,有真实场景和数据的垂直玩家更受青睐。

传统定位中,产业互联网公司是“交易平台”,价值在于撮合交易。但在Agent时代,产业互联网公司的新定位正在浮现——“产业AI应用平台”。十余年积累的场景、数据、业务流程和行业Know-how,正在从“成本中心”变成“核心资产”。从MaaS(模型即服务)到AaaS(Agent即服务)的演进中,产业互联网公司是关键的“最后一公里”交付者。

产业互联网公司的商业模式也将被重构:从售卖工具、收取订阅费,转向交付业务结果、按价值收费。

以NextB2B为例,据张晓坤介绍,未来将采用订阅费+Token消耗费+A2A交易服务费的混合模式。其中,A2A交易指企业Agent之间对话和交易,2026上半年带动GMV达到11.05亿元,同比增长73.2%。长远看,“Agent找Agent”将给B2B交易方式带来范式变化。

更重要的是,NextB2B展示了一种可能:产业公司可以成为AI落地的关键玩家,不仅仅是技术公司的客户,更是技术公司的行业合作方。大模型是底座,产业才是AI真正的考场。

找钢网先在内部使用AI提升效率,验证价值后将能力产品化并对外输出,从“内部提效”到“对外输出”的跨越并不容易,产品化能力、客户教育、商业模式验证、数据安全与信任,每一道都是坎。

贸易流通行业市场规模数十万亿元,数字化和AI化程度低,是一片蓝海。找钢网NextB2B模式一旦跑通,可复制到化工、汽配、货代等其他大宗商品贸易领域。

“AI的最后一公里,有可能就是咱们产业互联网的第一公里。”张晓坤精准总结:“B2B粗放式增长已彻底结束,做好精益化管理,用好AI,做好数据驱动,这个是必选题,不是一个可选题。”

而王东的判断则更进一步,在他看来,中国是全球最大的制造业国家,拥有巨大的企业间贸易市场,行业对效率提升和成本控制的追求永无止境,中国企业AI的普及可能走出一条不同于美国的路径——由入口级平台与垂直领域AI应用公司共同完成产业落地的“最后一公里”。

对于所有在产业里深耕的公司来说,AI时代不是终局,而是新的起点。

亿邦持续追踪报道该情报,如想了解更多与本文相关信息,请扫码关注作者微信。

本文首发于 亿邦动力 官方网站

文章来源:亿邦动力

广告
微信
朋友圈

FAQ回顾

企业级AI落地产业场景的核心难点是什么?

AI落地真实产业场景的核心难点并非技术不足,而是缺乏对应行业的场景、数据、流程与经验沉淀,通用AI不了解特定客户的真实需求、不懂行业生意策略,无法独立完成符合行业规则的全流程业务工作。

B2B贸易企业落地AI应用有什么独特优势?

B2B贸易企业属于对话驱动型组织,大量业务协同发生在微信、IM等对话场景中,天然积累了丰富的业务对话上下文,无需重复走先信息化再数字化的传统路径,可借助成熟AI智能体产品直接向AI原生阶段越级进化。

找钢网发布的NextB2B是一款什么产品?

NextB2B是找钢网联合腾讯WorkBuddy发布的面向B2B贸易场景的全流程AI Agent解决方案,将腾讯通用智能体能力与泛贸易流通行业经验结合,可为采购、销售、内勤、运营等岗位提供专属AI业务能力,降低企业AI化门槛。

2026年国内企业级AI Agent市场有哪些发展趋势?

2026年国内企业级AI Agent市场规模将突破480亿元,同比增速接近300%;行业投资逻辑从模型能力竞赛转向商业化落地验证,具备真实场景和数据积累的垂直玩家更受资本青睐,产业互联网公司将成为AI落地的关键交付角色。

这么好看,分享一下?

朋友圈 分享

APP内打开

赞 +1
+1
微信好友 朋友圈 新浪微博 QQ空间
关闭
收藏成功
发送
/140 0