杨林 国泰海通计算机分析师: 各位线上的机构朋友，大家晚上好。我是国泰海通看计算机的杨林那么非常感谢大家在这 个开工第一天，晚上 8 点接入我们今天的电话会。那么我们今天电话会主要内容，是汇报 一下这个讯测我的一些情况，因为我们这个也，之前也发了一份报告，应该是万得上唯一 一份这个线上的深度。很荣幸，就是我们挖掘到这家公司，因为从底部开始推荐，因为当 时市场也没有什么人关注，到现在加上春节不到一个，春节假期不到一个月时间，大概涨 幅应该有 80%。 然后我们认为公司还是很优质的，在这个实时数据处理这行业，从资本市场这个，也开始 是服务一些资管机构，到运营商，到能源，就他的下游行业，是在不断的丰富的。同时自 己的产品矩阵，也在不断的这个加大，就是大家能看到，基本上是从一，从数据的前期， 这个收集，这个治理呀，到后面的这个应用所以我们的标题也，这个也起得很直接，就是 打造成中国版的这个 Palantir 全生命周期的一些实时数据的这个服务。那么当下这个位置， 其实港股还是波动很大的，港股其实这几天分化很严重。 那我们觉得，确实还是有一定空间的，一个是马上，大家也应该也能看到，现在就有些新 闻，就是马上快入通了。那 ...

Core Insights - The article emphasizes the critical importance of data processing in the data analysis workflow, highlighting that it is the most time-consuming and essential step [3][12] - It outlines the six major steps involved in professional data processing, which include data cleaning, transformation, integration, calculation, sampling, and validation [4][5][6][7][8] Data Processing Value - Data processing is crucial as raw data often contains impurities such as duplicates, missing values, and inconsistencies, which can significantly affect the accuracy of analysis results [3] - The process of data cleaning and organization can lead to valuable insights, as anomalies and patterns may be discovered during this phase [3] Steps in Data Processing - The first step is data cleaning, which involves removing duplicates, handling missing values, addressing outliers, standardizing formats, and normalizing data [4] - The second step is data transformation, which converts raw data into the required format for analysis [4] - Data integration combines data from multiple sources, while data calculation derives new metrics from the original data [5] - Data sampling is necessary when dealing with large datasets, and data validation ensures the quality of processed data [5] Tools for Data Processing - Efficient data processing requires specialized tools, including Excel for small datasets, SQL for medium to large datasets, and advanced analytical tools like Python's Pandas and R's dplyr for robust data handling [6][7] - Business Intelligence (BI) tools such as Tableau and Power BI are used for visualizing processed data [8] Case Study - A case study illustrates how a team identified a fluctuation in conversion rates for a hotel room type due to multiple naming conventions on an OTA platform, which led to duplicate calculations. After standardizing the names, the conversion rate stabilized at approximately 3% [9] Common Pitfalls in Data Processing - The article identifies common pitfalls in data processing, such as over-processing, improper handling of data, neglecting validation, and lack of documentation [10][11] Professional Assurance from Teams - Professional teams provide standardized processing workflows, specialized technical capabilities, strict quality control measures, and comprehensive documentation to ensure data processing quality and reliability [12][13]

Core Viewpoint - Deutsche Bank has initiated coverage on XunCe Technology (3317.HK), positioning it as a leading player in China's real-time data infrastructure and analytics solutions, akin to "China's version of Palantir" [1][2] Group 1: Market Position and Business Model - XunCe is recognized as the "Data Agent first stock" and holds a leading position in the real-time data infrastructure and analytics market in China, particularly in asset management with an 11.6% market share [1] - The company has achieved full coverage of China's top ten asset management institutions, indicating its strong foothold in high-barrier industries [1] - XunCe's business model focuses on providing a "data operating system" rather than simple data display, integrating data collection, governance, computation, and analysis into clients' core business processes [3] Group 2: Growth Potential and Financial Projections - Deutsche Bank forecasts XunCe's revenue to grow from 632 million RMB in 2024 to 3.735 billion RMB in 2027, representing a compound annual growth rate (CAGR) of 81%, significantly surpassing industry averages [6][8] - The average revenue per user (ARPU) is expected to increase from 1.582 million RMB in 2022 to 2.724 million RMB in 2024, with a projected CAGR of 83% from 2024 to 2027 [7] - Revenue from industries outside asset management is anticipated to grow at a CAGR of 109% from 2024 to 2027, with telecommunications, urban management, and manufacturing identified as new growth drivers [8] Group 3: Profitability and Valuation - XunCe's gross margin exceeds 76%, significantly higher than the traditional IT outsourcing sector, providing a solid foundation for future profitability [8] - The company is expected to achieve adjusted net profit in 2026 with a net profit margin of 6.7%, increasing to 18.3% by 2027 [8] - Deutsche Bank's target price of 85 HKD implies over 60% upside potential from the current price of approximately 52 HKD, with a valuation that offers a significant margin of safety compared to global peers like Palantir and Snowflake [9][10]

Core Viewpoint - Xunce (3317.HK) is launching an IPO to raise up to HKD 1.238 billion, with shares priced between HKD 48 and HKD 55, indicating strong market interest in data infrastructure and analytics solutions [1] Group 1: IPO Details - The IPO will run from December 18 to December 23, offering 22.5 million H-shares, with 10% allocated for public sale in Hong Kong and the remainder for international placement [1] - The expected listing date for the shares is December 30 [1] - The entry fee for one board lot of 100 shares is HKD 5,555.47 [1] Group 2: Company Overview - Xunce is a provider of real-time data infrastructure and analytics solutions, catering to enterprises across various industries [1] - The company aims to utilize 80% of the net proceeds from the IPO for ongoing solution development and future R&D efforts [1] - 10% of the funds will be allocated to enhance marketing capabilities, while the remaining 10% will be used for working capital and other general business purposes [1]

Core Insights - Aurora Mobile is actively exploring the integration of emerging technologies, particularly quantum computing, into its existing business model, aiming for transformative changes in customer interaction and marketing technology [1] Data Processing and Insights: Quantum Acceleration - The company faces challenges with vast and complex data in its data processing and analysis business. Quantum computing's parallel processing capabilities significantly enhance the efficiency of analyzing customer behavior and preferences [2] - Quantum algorithms enable rapid identification of customer behavior patterns, allowing for the precise prediction of repurchase timing, thereby providing optimal marketing windows and improving data insight depth and accuracy [2] Precision Marketing and Recommendations: Quantum Empowerment - Precision marketing and personalized recommendations are core to the company's business. Quantum bits enrich customer profiling dimensions, creating more accurate and comprehensive customer images [3] - Enhanced collaborative filtering algorithms powered by quantum computing can aggregate vast amounts of information to tailor recommendations for customers, significantly increasing marketing conversion rates and customer satisfaction while optimizing marketing resource allocation [3] Intelligent Decision Support: Quantum Assistance - The complexity of marketing decision-making, with numerous variables, is addressed by the company's use of quantum computing for simulation capabilities. This allows for the construction of market simulation models based on market trends, customer dynamics, and competitor strategies [4] - Quantum simulations can test various promotional strategies, including different channels, timing, and promotional intensity, helping businesses make more informed and forward-looking marketing decisions [4]