With the rise of large-scale AI models, enterprise-level applications of artificial intelligence technology are flourishing, becoming one of the important means to upgrade contemporary business efficiency. However, for many companies, the huge data scale and algorithm research and development threshold are still like a natural chasm, and the industry is looking forward to new AI services that combine data and intelligence.
Tianyun Data is one of the few domestic technology companies that can provide both data infrastructure and a digitally native full-industry service chain. Adhering to the concept of "data intelligence", we are committed to the research and development and innovation of AI-Native databases and artificial intelligence software to promote enterprises to move toward digital nativeness and empower the digital transformation of industries.
In the second photosynthetic organization AI solution competition, MaximAI, Tianyun Data’s universal PaaS AI modeling platform product, performed well in building enterprise AI capabilities and successfully stood out, becoming the best among many participating solutions. and won the competition award
MaximAI is a platform that integrates machine learning, R&D and operations. We can use this platform to provide enterprises with comprehensive artificial intelligence support. Lu Hui, chief data scientist of Tianyun Data, said that MaximAI’s goal is to break through the usage restrictions of Hadoop/Spark, further reduce the cost of artificial intelligence applications, and realize the standardization, processization and batch production of artificial intelligence models, so that every enterprise can All can easily use artificial intelligence capabilities
Based on the user’s AI application needs, MaximAI brings multiple product values in scenario practice: for example, the platform uses a distributed architecture to support full data unbiased modeling of massive data, ensuring that the value of the data is deeply mined; those who are not familiar with programming Ordinary business personnel can also operate modeling, greatly reducing the threshold for modelers; the model can be streamlined and online by dragging and dropping, avoiding knowledge barriers and conflicts between different teams, and accelerating knowledge application; supporting statistics of data Functions such as analysis and visual exploration enable the modeling process to be traced back and the model to be reused.
For every aspect of the enterprise's value chain such as manufacturing, supply chain, marketing and sales, delivery and service, there is room for improvement using artificial intelligence. The application of MaximAI in business lines can reduce costs, improve efficiency, and meet the needs of business explosion. From a longer-term perspective, MaximAI can improve model performance, scalability, interpretability and accuracy, optimize resource allocation, improve model development efficiency and business agility, which undoubtedly maximizes the value of artificial intelligence.
"It is not easy to let AI enter thousands of industries and hundreds of enterprises." Qiao Wanglong, head of database research and development at Tianyun Data, said that Tianyun Data provides customers with a one-stop modeling process, which requires the platform to not only It is necessary to achieve full horizontal coverage in terms of functions, and at the same time, it is necessary to expand vertically in the industry field.
In the field of financial services, Tianyun Data has developed hundreds of quantitative operators at one time on the engineering platform based on the demand scenarios of various customers such as banks and securities firms. Subsequently, when providing services to other customers, these accumulated quantitative operators can help Tianyun data quickly enter new user scenarios and be reused again to realize the release of value
In the process of connecting user business scenarios, the compatibility issue between software and hardware has become more prominent. Especially for information and innovation fields such as finance, whether it can be adapted to domestic hardware bases with usable performance is the key to AI's advancement into thousands of industries. For Tianyun Data, Haiguang provides good support for the implementation of MaximAI.
"We have cooperated with Haiguang on projects as early as 2017 and will complete product compatibility adaptation in 2020." Lu Hui pointed out that in terms of compatibility and equipment performance, Haiguang's products are very user-friendly. high. "For example, in terms of server performance, according to feedback from front-end technicians, the performance of Haiguang servers is at least 10% higher than that of other servers."
It is worth mentioning that when Tianyun Data and Haiguang conducted product certification, they also received support and technical support from a number of software and hardware resources, including arranging for dedicated personnel to provide technical Q&A and special resource docking for Tianyun Data. Lu Hui said: "Because the database has very high requirements for hardware equipment, sometimes it is necessary to make some adjustments to the underlying technology so that it can run better on the equipment. Therefore, Haiguang's support for us is very important."
Currently, MaximAI can already support cloud-native AutoML, the training and release of multi-algorithm frameworks, multi-domestic chip support, and cloud-edge integrated application release and deployment. Tianyun Data has provided services to more than 50 Fortune 500 and listed companies, involving finance, technology, petroleum, electricity, biomedicine and other fields. "
As the winner of the second Photosynthetic Organization AI Solution Competition, I am very grateful that the competition provided a stage for MaximAI to showcase its results, and it also allowed Tianyun Data to deepen its cooperation with ecological partners such as Haiguang. "Lu Hui is very much looking forward to the next competition.
The above is the detailed content of Tianyun Data MaximAI creates ready-to-use AI capabilities. For more information, please follow other related articles on the PHP Chinese website!