On September 20, Lingyang Intelligent Technology (hereinafter referred to as Lingyang) held a special forum on "Data × AI: Intelligent Enterprise Services, New Momentum for Value Growth" at the 2024 Apsara Conference. Peng Xinyu, Vice President of Alibaba Group and CEO of Lingyang Intelligent Technology, released the annual intelligent strategy of products at the meeting: "(Algorithm + Computing Power + Data) x Scenario", emphasizing that enterprises must pay attention to scenarios, and only by deconstructing scenarios and reconstructing businesses can they truly embrace AI and bring breakthrough growth. Among them, in the field of data governance, Wang Sai, vice president of Lingyang, analyzed the advantages of Dataphin such as inclusiveness, high scalability, openness and intelligence, and launched Dataphin Agile Edition, semi-managed and DataAgent agents to help enterprises build "good data" that meets business requirements and lay a solid foundation for "good data".
Based on Alibaba's methodology and ten years of practice in retail, industrial manufacturing, Internet, finance and other industries, Dataphin provides one-stop big data capabilities for the whole life cycle of data collection, construction, management and use, and builds an enterprise-level data middle platform with reliable quality, convenient consumption, safe and economical production.
Traditional data governance faces two major challenges: first, privatized deployment, which can meet individual needs but is costly; The second is public cloud services, although the price is low, but it cannot provide personalized customization. In order to meet these challenges, Lingyang has upgraded Dataphin to a new generation of intelligent data governance products by combining AI technology this year, taking into account personalization and cost performance, providing semi-managed and agile service models, and equipped with DataAgent.
Wang Sai said: "In terms of data governance and data analysis, Lingyang's new changes this year are mainly reflected in two aspects: one is the innovation of architecture and capabilities, and the other is the improvement of AI technology. The new architecture aims to achieve inclusiveness, scalability, and openness to meet the data governance needs of enterprises of different sizes and stages, and complete the seamless upgrade of data governance. In terms of AI capabilities, product innovation and capability improvement have been achieved in intelligent data finding and intelligent data questioning. ”
Inclusive, low-cost start-up: the new agile version is launched
In the past, the investment scale of data warehouse projects was usually in the hundreds of millions of yuan. Dataphin Agile Edition meets the comprehensive data construction and governance requirements of small and medium-sized enterprises, allowing projects to grow from millions to hundreds of thousands, or even tens of thousands. This year, the core capabilities of Lingyang Dataphin have been comprehensively upgraded, so that customers can easily build and use Dataphin whether they are using MySQL or Oracle databases.
At the same time, Lingyang simplifies the use of full-stack data and product capabilities, and the construction of data storage and architecture becomes simpler, customers only need to purchase three standard servers on cloud services, and developers only need to know how to operate the database, and they can carry out a series of data construction and analysis work.
Evolvable upgrades: The engine and upgradeable capabilities continue to iterate
Dataphin's mature, complete, and scalable multi-modal data system architecture enables enterprises at different levels and at different stages of enterprise development to use and upgrade data governance tools in an evolutionary manner.
Small businesses can start at a lower cost, while midsize businesses can choose interactive, multiprocessor (MPP) type database engines, such as StarRocks or ClickHouse, for enhanced performance and scalability. Lingyang supports database engines such as Hologres, Lindorm, and Impala, ensuring that they can be seamlessly integrated with Dataphin. For large enterprises, Lingyang provides the combination of big data engine and database, as well as the support of the lakehouse architecture, to ensure seamless upgrade and expansion from small to large enterprises.
Launched the semi-managed model on the cloud to make personalization more cost-effective
Previously, enterprises needed a few days to deploy, but the semi-managed model on the cloud only took 1 hour, and they could enjoy villa-level independent deployment products and services on the public cloud. This model provides significant savings in initial usage costs, and users only pay for the services they use on a pay-as-you-go basis, which is expected to save about a quarter of the cost compared to standalone deployments. In addition, users can enjoy a completely independent and autonomous space, while also meeting the high standards of data security management requirements of current Chinese enterprises.
In addition, in large group companies, the harmonization and standardization of data across multiple organizations, as well as the ownership and control of data across teams, are key requirements. To this end, Dataphin provides multi-tenant management capabilities, allowing enterprises to build multiple tenants within the enterprise to achieve distributed storage and management of data, while ensuring the unification and delivery of data standards. Enhance collaboration and communication between tenants while providing clear control. This approach aligns with One Data's philosophy of building a unified data environment.
For situations with high security requirements, Dataphin can achieve one-click production, ensure that the physical environment and production environment can be quickly switched, and ensure data security and controllability through tenant isolation and deployment in different environments.
At the same time, Lingyang provides developers with highly convenient product openness capabilities, allowing the development of enterprise-defined products and functions based on products. Whether it is enterprise internal login, embedded analysis, API development based on raw data, or batch operation, Lingyang provides an easy-to-use open interface.
Dataphin· DataAgent: Lowers the threshold for data governance and allows everyone to have their own agent
How can business people easily turn data into practical applications when data inspiration comes to them, and what are the obstacles in this process? Is it a lack of data thinking, programming skills, or a lack of understanding of data? In the past, this gap needed to be bridged by data experts. Wang Sai emphasized that Lingyang is to build an AI-based DataAgent to achieve direct and effective communication between business personnel and data, thereby significantly improving efficiency.
With the addition of DataAgent, users can build an enterprise-specific data asset agent in just three steps. First, you can build a vector database of selected data assets and tables with one click. On top of that, users can customize the arrangement and layout of the DataAgent. Finally, chatbots can be published on the data asset platform with one click to enable dialogue and communication.
The underlying layer of the platform is Dataphin's global assets and Alibaba Cloud's Tongyi Qianwen model, and enterprises can choose different data models according to their needs. This is not just a set of solutions, but a platform-level capability that allows enterprises to build agents based on their own knowledge and capabilities.
In the case of data cataloging and retrieval, the data searched in the past was only related to keywords and could not fully express how the data was used. Now, with the Asset Q&A capability, users can easily associate interested and related information directly into business processes.
Under the general trend of AI technology, Lingyang is committed to making data governance and data analysis more convenient and inclusive, so that all organizations can use advanced data tools to build and use data, so as to find ways to successfully integrate data and AI to reconstruct business scenarios.