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

With the rise of AI large model tuyere, the enterprise application of artificial intelligence technology blossoms and becomes one of the important means to upgrade contemporary business efficiency. However, for many enterprises, the huge data scale and algorithm research and development threshold is still the same as natural cutting, the industry looks forward to new AI services with data and intelligence.

Tianyun data is a rare technology company in China that can provide data infrastructure and digital native industry-wide service chain at the same time. Adhering to the concept of "data + intelligence", we are committed to the research and development and innovation of AI-Native database and artificial intelligence software, so as to promote enterprises to move forward to digital native and empower industry digital transformation.

In the second photosynthetic Organization AI solution Competition, Tianyun data's general PaaS AI modeling platform product-MaximAI, handed over a new answer paper for enterprise AI capacity building, and stood out among many participating schemes and entered the award list of the competition.

"MaximAI is an integrated platform covering machine learning, R & D and operation, and we can use this platform to provide a full range of AI empowerment for enterprises." Lu Hui, chief data scientist of Tianyun data, said that MaximAI aims to break through the bottleneck of the use of Hadoop / Spark, further reduce the landing cost of AI, and realize the standardization, flow and batch production of AI models, so that every enterprise has the ability to use AI out of the box.

Based on users'AI application requirements, MaximAI brings multiple product values in scene practice: for example, the platform uses distributed architecture to support unbiased modeling of massive data to ensure that the data value can be deeply mined; ordinary business personnel who are not familiar with programming can also operate modeling, greatly reducing the threshold for modelers By dragging and dragging, the model can be pipelined and online, the knowledge barriers and contradictions between different teams can be avoided, the knowledge application can be accelerated, and the functions such as statistical analysis and visual exploration of data can be supported, and the modeling process can be backtracked and the model can be reused.

From the value chains of manufacturing, supply chain, marketing and sales, delivery and service, there is room for improvement by using AI in every link, while MaximAI can reduce costs and increase efficiency on the business line to meet the needs of business outbreaks. In the longer term, MaximAI can improve the performance, scalability, interpretability and accuracy of the model, optimize the allocation of resources, improve the efficiency of model development and business agility, which undoubtedly maximizes the value of AI.

"it's not easy to get AI into a hundred enterprises." Qiao Wanglong, head of research and development of Tianyun data database, said that Tianyun data provides customers with an one-stop modeling process, which requires the platform not only to achieve horizontal full coverage in function, but also to expand vertically in the industry field.

In the service field, Tianyun data has developed hundreds of quantitative operators at one time on the engineering platform according to various customer demand scenarios such as banks and securities firms. Next, when serving other customers, these accumulated quantization operators can help Tianyun data quickly cut into the scenario of new users, reuse again and release value.

In the process of connecting user business scenarios, the compatibility between software and hardware has become more prominent. Especially for the field of innovation and innovation, whether it can be adapted to the domestic hardware base with available performance is the key for AI to move towards thousands of industries. For Tianyun data, Haiguang provides a good support for the landing of MaximAI.

"We cooperated with Haiguang on the project as early as 2017 and completed the compatibility adaptation of the product in 2020." Lu Hui pointed out that Haiguang products are very friendly in terms of compatibility and equipment performance. "for example, in terms of server performance, according to feedback from front-end technicians, Haiguang server performance is at least 10% better than other servers."

It is worth mentioning that Tianyun data and Haiguang product certification, but also received a number of software and hardware resources support and technical support, including the arrangement of special personnel for Tianyun data to provide technical answers, special resources docking, and so on. " Because the database has very high requirements for hardware equipment, sometimes it is necessary to make some changes in the underlying technology in order to run better on this device. So Haiguang's support for us is very important. " LV Hui said.

At present, MaximAI can support cloud native AutoML, multi-algorithm framework training and release, multi-domestic chip support, cloud-side integrated application release and deployment. Tianyun data has provided services to more than 50 + top 500 and listed companies in many fields, such as science and technology, oil, electricity, biomedicine and so on. "

As the award winner of the second photosynthetic Organization AI Solutions Competition, I am very grateful to the competition for providing a stage for MaximAI to showcase its results and for deepening the cooperation between Tianyun data and Haiguang and other ecological partners. "Lu Hui is very much looking forward to the next event.

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.