Shulou(Shulou.com)11/24 Report--

On September 19, Volcano engine announced at its V-Tech data-driven Technology Summit that VeDI, a mathematical platform for volcano engines, has launched a "AI Assistant" to help enterprises improve the efficiency of data processing and query analysis by accessing large models of artificial intelligence. Even operators who do not know how to write code can do a good job of fetching, viewing, and attribution analysis of business operation data by talking to the big model. At present, the VeDI-related data products have started the invitation test.

▲ figure: Tan Dai, president of volcano engine, believes that data flywheel will be an important direction for the application of large models in the enterprise market. He said that the big model lowers the threshold for enterprises to mine data value and more efficiently build data flywheels with data consumption as the core. "AI-enhanced data flywheels will become a new paradigm for enterprises to do a good job of data-driven."

Data full-link AI application

According to reports, big data R & D governance suite DataLeap of volcano engine VeDI and Intelligent data Insight DataWind have been equipped with AI assistant to cover the full link scenario of data production and consumption, including data asset query, development, operation and analysis insight.

The query and development of data assets is the basis of data consumption. In the past, the retrieval and development of data assets relied heavily on professionals, while large models greatly lowered the threshold. Non-R & D personnel use DataLeap- to find numbers efficiently and accurately through conversational query, realizing the first step of employee self-help data consumption; based on DataLeap- development assistants, using natural language, they can automatically generate and optimize SQL codes, as well as dialogic consultation on the use of SQL, to break technical barriers and make data development simple, efficient and convenient.

In the process of data analysis, a series of business explorations such as data visualization query and analysis can be completed by using DataWind- analysis assistant and enterprise employees talking with large models, which can solve the pain points that require a lot of professional knowledge in the past analysis insight, and shorten the data analysis cycle. At the same time, DataWind is also connected to Flying Book and other office collaborative tools, users through IM message subscription, dialogue, more extension analysis, to achieve flexible analysis anytime, anywhere.

"VeDI's two products not only lower the threshold for non-professionals' data consumption, but also liberate professionals to focus on the needs of complex scenarios and improve R & D production efficiency and code quality," Tan said.

Get App co-founder and CEO to think that data-driven is essentially driving the formation of consensus within the enterprise. " If 'data consumption' is difficult and complex, more than half of the people will not look at it for various reasons, "says Buhua." Volcano engines push all kinds of data Kanban in front of everyone, causing discussion and forming a consensus on action. The acceleration of the first kick is very important if the data flywheel is to turn. "

Change the way to explore the value of data

▲ diagram: the application of Luo Xuan Da model in charge of byte jump data platform in data products has achieved initial results in byte internal business practice. Luo Xuan, head of the byte jump data platform, showed a case of linkage analysis of DataLeap and DataWind in e-commerce scenarios.

In the past, e-commerce operators needed to turn to R & D personnel to pick up the number in DataLeap. Now, e-commerce operators who do not know how to write codes can directly ask: "which tables should be used for the operation of the live broadcast room of good things in the last 7 days?" According to the knowledge base of the business, DetaLeap recommends tables related to operating conditions and explains the corresponding data dimensions of each table.

Operators can allow DataLeap to automatically generate data requirement codes corresponding to business conditions, such as order sales by city, or traffic between live broadcasts in different periods. Operators can also question the meaning of the code; they can also parse the generated code with one click, call the SQL tool to check the table, click to confirm AI automatic repair, and further optimize data assets.

With data assets, operators create datasets through DataWind drag-and-drop, and use natural language to define the logic of different fields, such as directly looking up the data of "big shot live time". After checking, the operators will be able to visually analyze and explore.

DataWind provides automatic AI analysis to facilitate further exploration of the reasons behind the chart. For example, in the visual charts such as "Top category of studio sales", operators achieve business insight through further attribution in the form of dialogue based on the results of AI analysis.

"currently, 80% of employees within byte jump can directly use data products, and data assets that can be managed and operated cover 80% of daily analysis scenarios." Luo Xuan said that with the blessing of the ability of large models, data consumption has become universal, and the business is also more data-driven.

Help enterprises to land data-driven

In April this year, based on the practical experience of byte-beating data-driven, the volcano engine released a new digital intelligence upgrade model "data flywheel" for enterprises. the core element is to do a good job in data consumption. The Volcano engine Digital Intelligence platform uses the latest AI technology to further reduce the threshold for data consumption.

As a new type of production factor, data is supporting the digital intelligence transformation of enterprises. Many enterprises have more digital construction, but can not release the data value better. There are some problems, such as high cost of data construction and management, high threshold for the use of data products, and low value of data assets. The reason is that there is no two-way benign drive between business and data.

▲ figure: volcano engine data flywheel data flywheel, through data consumption to achieve scientific insight, efficient decision-making, agile action, resulting in increased business value; but also through frequent data consumption and business benefits, targeted construction of high-quality, low-cost data assets to better support business applications. Enterprise data flow is fully integrated into business flow to achieve the flywheel effect of data assets and business applications, and stimulate enterprise vitality.

With data consumption as the driving force, it has brought real benefits to the partners of the volcanic engine. Li Yifan, director of the new business development division of Boxi Home Appliances, said at the data-driven Technology Summit that based on the volcano engine digital intelligence platform, Boxi Home Appliances has been able to achieve more convenient and frequent data consumption in digital marketing and user operation. Taking the operation of "Kumgang bit" (the functional entrance navigation area below Banner) as an example, the click conversion rate of Bosi Home Appliances operation bit has increased by more than 50%.

Liu Hang, head of digital Jetway, a car brand owned by Chery Holdings, revealed that operators' digital analysis on the volcano engine digital platform is more independent. The APP operation data analysis scenario has been shortened from the past 7 days to the current second response, and many accurate hierarchical operation scenarios have been expanded from 0 to 1. For example, in maintenance package sales scenarios, SMS messages achieve more than 90% opening rate.

Tan Dai said that building a data flywheel centered on data consumption will be the inevitable trend of enterprise digital construction. "the volcano engine VeDI will be further upgraded to continuously lower the threshold of data consumption through AI + data flywheel, fully release the data value of enterprises, and help enterprises better gain growth momentum from the data."

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.