Jinan Social Security Center Deploys 79 AI Digital Worker Scenarios Covering 4.28 Million Operations

Title

Jinan Social Security Center Deploys 79 AI Digital Worker Scenarios Covering 4.28 Million Operations

Content

At the Jinan Social Security Center, a special team of digital workers has been deployed to handle social insurance business 24/7 without breaks. These AI-powered digital workers leverage Alibaba Tongyi Qianwen 3 and DeepSeek large language models in combination with RPA robotic process automation, OCR image recognition, and multimodal interaction technologies to build an intelligent service system covering the entire business chain. For high-frequency operations such as enterprise pension transfer, social security card application review, and insurance registration, the digital workers automatically capture information and coordinate across systems for fully automated processing. For enterprise pension transfer specifically, a process that previously required staff to switch between multiple systems is now handled automatically, with processing time significantly reduced and citizens no longer needing to wait in lines. The digital workers are equipped with advanced OCR technology that can instantly read image information and intelligently determine material validity and compliance. Where two staff members previously needed a full day to review materials, the digital workers can complete the same volume of review in moments, increasing social security card application review efficiency by over 80% and eliminating the inefficiency of manual inspection. The digital workers leave complete traces throughout the entire process with operation logs, screenshots, and screen recordings for traceability and auditability. As of now, the Jinan Social Security digital workers have launched 79 application scenarios covering pension insurance, work injury insurance, and all insurance types, spanning the entire process of business handling, review, and risk prevention. The system has handled 4.28 million various operations with an automatic review rate of 87%, saving over 76,000 hours of handling time. The project has received recognition including the 2024 Shandong Province Information Technology Application Innovation Solution award and the 2025 Shandong Division Data Elements Application Competition second prize, and was featured on CCTV News, making it a demonstration case for national government AI applications.