Customer Service Automation · Enterprise Knowledge Search
Hualvwang needed a way to help ordinary users describe legal issues more precisely, retrieve relevant legal knowledge faster and scale personalized legal-service triage across a very large user base.
Reported impact
Hualvwang improved legal-intent discovery, accelerated access to relevant knowledge and services and created a higher-volume, better-targeted legal-service funnel on top of its content assets.
Hualvwang used Baidu AI Cloud's Qianfan large-model platform to build two tightly related legal-service features on top of its huge legal corpus. The first is the 'recommended questions' flow, which uses Baidu's generation, understanding, logic and memory capabilities to refine vague user requests into clearer legal-service intents and then steer users toward the right area of law, knowledge and follow-up services. The second is the 'Xiaolvxia' legal adviser, which can generate legal-opinion content from real-time consultations and follow-up questionnaire responses. Under the hood, the platform relies on more than two decades of Hualvwang knowledge assets, more than 900 million legal-knowledge entries and large-scale consultation data to ground model outputs in the site's own domain material. This turns the model from a generic chat interface into a two-sided platform asset: it improves user-side discovery and simultaneously helps the business assess lead value, connect people with appropriate legal experts and strengthen the site's content ecosystem.
User legal questions enter the recommended-questions workflow, which uses domain knowledge and model reasoning to refine intent and retrieve relevant legal content; follow-up answers then feed the Xiaolvxia adviser, which generates legal opinions and assessment forms for downstream service routing.
Daily model invocations exceeded 10 million at peak (original: 文心大模型日均调用量峰值超过1000万).
Hualvwang used Baidu AI Cloud's Qianfan platform and ERNIE to build the 'recommended questions' feature from zero and create the 'Xiaolvxia' legal adviser (original: 基于百度智能云千帆大模型平台调用百度文心大模型，聚焦法律咨询场景，帮助客户从0到1搭建「推荐问」功能……同时，大模型助力客户打造「小律侠」法律顾问).
The model can help connect users to the right legal experts with 97% accuracy (original: 帮助对接到该细分领域的法律专家，准确率达到97%).
The platform has more than 900 million legal-knowledge entries and more than 20 years of high-quality legal support (original: 平台沉淀优质法律知识超9亿条……基于华律网超大规模的用户数据以及超过20年沉淀的优质法律支持).
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
Baidu AI Cloud Sample May 15, 2025 Wuhan Tong Technology Co., Ltd. (武汉通科技有限公司) Wuhan Tong Technology needed to help severely understaffed grassroots government teams serve large resident populations, answer policy questions more accurately and automate labor-intensive work-order handling across… China has 477,800 administrative villages and 116,000 communities, with only 3-8 cadres typically serving each… Government Knowledge Base · Government Service Copilot · Qianfan AppBuilder 
ServiceNow Sample Modified · Jul 30, 2026 ServiceNow ServiceNow wanted to close the gap between consumer-grade self-service and the fragmented enterprise support experience, where customers and employees still had to navigate multiple portals, unclear instructions, and… 2.3 million hours saved in 2025 AI Search · AI agents · Knowledge Management 
NVIDIA Sample Jul 7, 2026 Siam AI Corporation Co., Ltd. (Siam.AI Cloud) Thailand tourism information was fragmented across multiple websites, difficult to verify, often outdated, and hard for visitors to access across languages, creating poor experiences and inefficient support workflows. NVIDIA AI Enterprise · NVIDIA NIM · NVIDIA NeMo 