FAW Group, Alibaba Cloud launch GPT-BI large model-based data-driven application

Beijing (Gasgoo)- On January 22, FAW Group, a major Chinese automaker, in collaboration with Alibaba Cloud, introduced the groundbreaking GPT-BI (Business Intelligence) application, further boosting FAW Group's digital transformation. 

Photo credit: Alibaba Cloud

This application, which is based on Alibaba Cloud’s Tongyi Qianwen (Qwen) large model, is capable of receiving natural language queries, leverages enterprise data to generate analytical charts with an impressive accuracy rate of nearly 90%. Notably, unlike traditional BI’s fixed queries, the GPT-BI allows for unlimited question combinations, enabling real-time data penetration and achieving instant insights through interactive questioning.

Photo credit: Alibaba Cloud

Leveraging Alibaba Cloud’s capabilities and 468 governed indicators, the GPT-BI underwent continuous refinement, boosting its overall accuracy from an initial 3.2% to an outstanding 90%, surpassing average human governance levels. FAW's large-scale data model seamlessly transitions core domain data governance into comprehensive enterprise-wide data governance.

The GPT-BI significantly reduces BI analysis delivery cycles and offers exhaustive coverage of enterprise metrics, models, and reports within limited domains. The model interprets user queries, decodes decision variables, generates SQL data queries, matches real-time enterprise data, and autonomously formulates optimal decision solutions, meeting users' flexible and intelligent data demands. This achieves the revolutionary concept of "questioning equals insight" based on dynamic factors and real-time data.

For instance, when asked, "Why is the production volume of a certain vehicle model below expectations?" the GPT-BI compares expected and actual production volumes, conducts a thorough analysis beyond explicit variables (such as equipment-related anomalies causing a 20-minute production halt or abnormal quality in certain model parts), delving into all relevant factors (e.g., fluctuations in raw material supply, energy consumption, and supply stability). The model identifies the most significant correlations and generates visual reports.

FAW Group, through GPT-BI and real-world business scenarios, plans to utilize its extensive high-quality data assets to develop vertical large-scale models for R&D, manufacturing, after-sales service, and more. Concurrently, Alibaba Cloud is driving the application of the Qwen large-scale model in diverse industries such as aviation and automotive, contributing to the development of a collaborative industry ecosystem.