文 丨 《财之道》 周文猛 今年初，英伟达创始人黄仁勋提出了AI发展的四个阶段：感知AI（Perception AI）→生成式AI （Generative AI）→代理式AI（Agentic AI）→物理AI（Physical AI）。并进一步指出："代理式AI才是 真正的智能觉醒"。 作为英伟达合作伙伴网络（NPN）成员之一，从感知AI时代就投身AI创业浪潮当中的视达创始人、 CEO吴一黎，对这一提法深表认同。 近日，他在与新浪财经《财之道》栏目对话中直言，"我们今天其实已经进入Agentic AI时代，感知AI时 代的AI更多是一个工具——只能帮你完成某一个功能，但今天的AI已经是一个同事和伙伴，有了'超级 智能'，能独立自主地完成工作。" "工具是最终完成这个工作的对象还得是我，但Agentic AI自己有一套思考逻辑，能全部做完，并直接给 你结果。"吴一黎以视达创业之初便锚定的AI零售行业为例指出，"今天的AI在零售领域，会由过去一个 督导管8家便利店，变成800个督导管1个店，而且成本可能不到原来的1/100。" 从2016年"卖掉北京房产"自筹资金创业，到如今将公司业务拓展至30个国家和地区，吴一 ...

Group 1 - The core viewpoint of the news is that the Chinese government is promoting consumption upgrades to drive industrial upgrades, which aligns with the strategic direction of the company, Multi-Point Intelligence [1] - Multi-Point Intelligence has recently repurchased a total of 590,000 shares, investing approximately HKD 431.57 million, signaling confidence in its value and industry prospects [1] - The company has received authorization from shareholders to repurchase up to 10% of its total shares, with funding sourced entirely from its own capital [1] Group 2 - Multi-Point Intelligence is a key player in retail digital transformation, leveraging AI as a core driver to innovate and optimize its product matrix [2] - The company has established partnerships with 438 leading retail clients both domestically and internationally by the first half of 2025, enhancing its market influence [2] - The recent government policy aligns with the company's strengths in supply chain management and integrated online-offline operations, creating a favorable environment for its digital transformation services [2] Group 3 - According to a report from招商证券, the digitalization rate in the retail industry in China and Asia is relatively low, indicating a growing demand for lean operations [3] - Multi-Point Intelligence, as the leading provider of retail cloud and digital solutions in China, is well-positioned to capitalize on the significant growth potential in the market [3] - The company is rated as "overweight" due to its strong customer base and ongoing expansion into overseas markets [3]

Core Viewpoint - The company believes its current market value is below its intrinsic value and proposes a share buyback to enhance shareholder value and demonstrate confidence in long-term growth [1][2] Group 1: Business Operations and Financial Performance - The company maintains a positive growth trajectory as a leading AI retail solution provider, focusing on technological innovation and retail process needs [1] - As of the first half of 2025, the company has established partnerships with 438 leading retail clients globally, enhancing its customer network's quality and coverage [1] - The company achieved a net revenue retention rate of 107% by the first half of 2025, solidifying its industry-leading position [2] - The company has a strong cash reserve, with cash and cash equivalents amounting to RMB 927 million as of October 31, 2025, ensuring the feasibility of the share buyback without impacting normal operations [2] Group 2: Strategic Initiatives - The company is committed to continuous product matrix improvement and has launched new AI features to address industry pain points [1] - The board believes that the share buyback reflects a strong recognition of the company's intrinsic value and confidence in its long-term business prospects [2] - The share buyback is expected to enhance value recognition and provide stable long-term returns for shareholders, aligning with the overall interests of the company and its shareholders [2]

Core Insights - The retail industry's competitive paradigm has fundamentally shifted, moving away from traditional methods of increasing manpower and budgets to a focus on building an efficient data-driven decision-making loop [2] Group 1: Retail AI Evolution - The evolution of retail AI can be categorized into three waves: - The first wave is digital record-keeping, which involves moving business operations online for traceability [3] - The second wave is automated recognition, utilizing CV/OCR technology to address visibility issues [5] - The third wave focuses on reasoning and closed-loop systems, transitioning from mere observation to understanding and execution [5] Group 2: Operational Efficiency - The current challenges in terminal execution include out-of-stock (OSA) issues, inconsistent price tags, and improper displays, which cannot be resolved solely through traditional computer vision techniques [5] - A breakthrough is achieved through the integration of visual data, vertical knowledge, and reasoning [5] Group 3: AI Implementation and Impact - The company emphasizes a methodology that ensures AI capabilities directly impact core business metrics and profit margins [7] - A forward-looking concept introduced is that retail execution is becoming software-driven, with KPIs transforming into APIs, allowing for automated, trackable, and reusable execution instructions [7] - This transformation leads to a programmable and closed-loop retail operation model, significantly enhancing efficiency [7] Group 4: Results and Future Vision - The implementation of AI-driven end-to-end closed loops can reduce manual review workloads by 60-80%, improve promotional execution accuracy by 8-12 percentage points, and enhance OSA by 2-4 percentage points while reducing store inspection time by 40% [9] - The company's mission is to empower the industry by converting cutting-edge AI technology into user-friendly solutions that deliver tangible value, emphasizing that AI's true worth lies in its ability to improve margins and efficiency [9]

Core Viewpoint - The company reported a revenue of 1.078 billion yuan for the first half of 2025, reflecting a year-on-year growth of 14.8%, although the growth rate has slowed compared to 2024 [1][2] Revenue Performance - AI retail solutions generated revenue of 490 million yuan, up 16.2% year-on-year, while AI retail value-added services brought in 590 million yuan, increasing by 15.0% year-on-year [1] - The revenue growth rate for AI value-added services significantly decreased due to a drop in labor outsourcing income and costs, but the gross margin improved to 13.7% from 9.0% in the previous year [1][2] Profitability - The company achieved a net profit of 62.174 million yuan, marking a turnaround from losses, with adjusted profit from continuing operations at 77.012 million yuan, a 152.5% increase year-on-year [2] - The net profit margin improved due to a reduction in R&D and administrative expense ratios, with R&D expense ratio at 17.6%, down 4.1 percentage points, and administrative expense ratio at 12.9%, down 1.3 percentage points [2] Strategic Initiatives - The company is expanding domestic and international collaborations, partnering with notable clients like Pang Donglai and Shanghai Sugar and Wine Group, which is expected to enhance operational efficiency and customer acquisition [2] - The implementation of AI productivity tools has led to significant cost reductions and efficiency improvements, contributing to the company's profitability [2] Future Outlook - The company maintains revenue forecasts of 2.17 billion yuan, 2.54 billion yuan, and 2.96 billion yuan for 2025-2027, respectively, while net profit forecasts have been adjusted upwards to 130 million yuan, 280 million yuan, and 420 million yuan for the same period [3] - The current price-to-sales ratios are projected at 4.4x, 3.7x, and 3.2x for 2025-2027, with a "buy" rating maintained [3]

Investment Rating - The report maintains a "Buy" rating for the company, indicating an expected investment return exceeding 15% over the next 6-12 months [5][23]. Core Insights - The company achieved a revenue of 1.078 billion CNY in H1 2025, reflecting a year-on-year growth of 14.8%, although the growth rate has slowed compared to 2024 [1]. - The revenue from AI retail solutions reached 490 million CNY, growing by 16.2%, while AI retail value-added services generated 590 million CNY, with a growth of 15.0% [1]. - The company focuses on high-quality growth, with an emphasis on cost reduction and efficiency improvement through AI productivity tools, leading to a significant decrease in expense ratios and a return to profitability [2][3]. Revenue and Profitability - The company reported a net profit of 62.17 million CNY in H1 2025, marking a turnaround from previous losses, with an adjusted profit of 77.01 million CNY, up 152.5% year-on-year [3]. - The gross margin for AI value-added services improved to 13.7% from 9.0% in H1 2024, while the gross margin for core AI retail solutions decreased to 68.3% from 76.8% due to increased costs associated with project deliveries [2]. Future Projections - Revenue forecasts for 2025-2027 are set at 2.169 billion CNY, 2.542 billion CNY, and 2.960 billion CNY, respectively, with corresponding growth rates of 16.7%, 17.2%, and 16.4% [4]. - The net profit projections for the same period have been revised upwards to 130 million CNY, 280 million CNY, and 419 million CNY, reflecting increases of 6.5%, 3.5%, and 2.7% respectively [4]. Market Position and Strategy - The company is actively expanding its domestic and international partnerships, including collaborations with notable clients such as Shanghai Sugar and Wine Group and expanding its cooperation with retailers like Pinduoduo and Lawson [3]. - The focus on digital transformation in the retail sector is expected to drive accelerated customer order growth, particularly through the implementation of successful pilot programs [4].