Reported impact
Business outcome
The deployment gave local governments a 24/7 resident assistant and automated work-order handling, helping Wuhan Tong deliver smarter public-service software while improving frontline efficiency, speeding resident responses and raising satisfaction.
- Customer Experience
- Productivity Improvement
- Scalability
Solution
Wuhan Tong Technology Co., Ltd. used Baidu AI Cloud's Qianfan stack to build a grassroots government assistant spanning district, street and community operations. On the resident side, the team used Qianfan AppBuilder roles, follow-up prompts, speech recognition and Online-Search to expose consultation through WeChat groups, personal assistants, mini-programs, official accounts and H5 endpoints. Government policies and files were consolidated into a knowledge base so the assistant could answer common questions locally and fall back to controlled online search when standard content was missing. On the staff side, AppBuilder, ModelBuilder, NLP, Workflow and GBI convert patrol notes, dispute visits and daily resident conversations into structured forms and work orders, then route, review, feed back, evaluate and close cases against government procedures. The implementation also ingests crawled governance information to support intelligent analysis and category-based dispatching.
Data flow
Resident questions enter through WeChat groups, mini-programs, official accounts and H5 clients into the Resident Assistant on AppBuilder. The system answers from the local government knowledge base first, then uses controlled online search when standard content is missing. Staff-side text and voice inputs pass through ModelBuilder and NLP into normalized forms and work orders, which Workflow routes through review, feedback, evaluation and closure.