Ant Shield releases "knowledge interactive modeling engine" for the real industry, customizing AI risk control models in as fast as 10 minutes

Traditional industries face many difficulties in the development of intelligence, such as the late start of digitalization, scattered and sparse data, high professional barriers, and reliance on "old masters" for industry knowledge. At the 2023 Yunqi Conference, Ant Group's security technology brand Ant Shield released the "Knowledge Interactive Modeling Engine", which organically combines physical industry knowledge with AI models to provide enterprises with customized and controllable solutions in as fast as 10 minutes. Vertical field AI risk control engine. At present, this engine has been the first to be tested in the commodity industry

## The 20th National Congress report elevated “ensuring the security of important industrial and supply chains” to the level of national security. Taking upstream and downstream industry collaboration as an example, vicious events such as breach of contract, runaways, and thunderstorms not only bring financial losses to enterprises, but also affect inter-industry collaboration trust and hinder the healthy development of industries. However, traditional supply chain risk control methods based on human flesh research and manual experience analysis can no longer meet the current high requirements for risk timeliness and accuracy.

Based on Ant's more than ten years of practical experience in risk control, Ant Shield has developed a one-stop enterprise risk control platform that focuses on managing "corporate business risks" ". Manual due diligence that originally took several days to complete can now be completed in the shortest time, helping the business reduce risks and improve efficiency

Ant Shield Product Leader Li Sha said in her sharing, "The traditional credit risk control model is based on general financial knowledge and fails to fully integrate industry data and knowledge, and cannot be directly reused in industrial risk control. Each industry has different access qualifications, business Form, data latitude, policy background, upstream and downstream risk transmission relationships, etc. Therefore, each industry and each form of cooperation should have different risk judgment standards to achieve more accurate risk control." In the field of commodities, Ant Shield It has jointly built dozens of vertical industry models such as steel plants, cement plants, traders, and construction industries with leading industrial enterprises such as Helian Group and China Industrial Zhongda Group.

#People think “data and Artificial intelligence driven by "knowledge two wheels" is the next climax of artificial intelligence and will accelerate the application of artificial intelligence in real industries. It is understood that Ant Shield has established a compliant and credible data pipeline for enterprises, integrating and upgrading the "enterprise map" of customer-merchant-shareholder relationships and the "industry map" of industrial chain collaboration, creating a network that carries 6,000 industries and nearly 100 million enterprises. , an "industrial map" with 2 billion relationships, drawing supply and demand relationships, production relationships, risk transmission relationships, etc. between enterprises from macro and micro perspectives. At the risk control application level, Ant Shield injects the 40,000 risk factors accumulated by Ant into the industry map to paint a more comprehensive risk portrait of merchants from the aspects of merchants' operating capabilities, credit risks, public opinion risks, etc.

The "knowledge interactive modeling engine" released by Ant Shield this time is a new function of its industrial risk control platform. Based on the universal algorithm base, traditional enterprises can inject industry experience through interaction with AI, and build a personalized risk control engine for vertical industries in as fast as 10 minutes. In addition, this engine solves the problem of insufficient risk sample data of the enterprise itself, and realizes automated sample expansion and label expansion through knowledge interaction enhancement technology, which helps to quickly form a customized model that conforms to the enterprise's risk perception. Zhang Peng, head of Ant Shield’s data algorithm, said, “This innovative attempt can make modeling work flexible and efficient. It will also help Ant Shield expand into more vertical industries and allow AI risk control to deeply serve the real industry.”

The above is the detailed content of Ant Shield releases "knowledge interactive modeling engine" for the real industry, customizing AI risk control models in as fast as 10 minutes. For more information, please follow other related articles on the PHP Chinese website!