Customer Service Automation
China Unicom needed to cut the cost of serving customers across hotline and online channels, break down siloed service systems, and modernize aging provincial deployments without compromising support quality for a very large subscriber base.
Reported impact
China Unicom gained a centralized intelligent customer-service stack that improves cross-channel consistency, supports lower-cost autonomous operations, and provides a scalable foundation for wider national rollout and service upgrades.
China Unicom adopted Baidu AI Cloud's ABC smart customer-service solution to unify the 10010 hotline, internet service channels, and the broader intelligent-customer-service operating model behind one centralized platform. Keyue Intelligent Customer Service Enterprise Edition delivers the customer-facing service layer, while the Keyue voice engine handles high-accuracy speech recognition and semantic interpretation. Responses are grounded in a structured enterprise knowledge graph, and a supporting AI operations platform lets China Unicom train models, configure flows, test, publish, analyze, and optimize services without rebuilding the stack each time. The deployment also uses a container-based microservice architecture so capacity can scale elastically under high concurrency. Together, those layers give China Unicom a practical way to replace fragmented channel tools with a shared AI foundation that supports nationwide rollout and lower-cost long-term operations.
Requests from the 10010 hotline and online service channels enter the Keyue platform, where the voice engine converts speech and interprets semantics. The dialogue layer then queries the enterprise knowledge graph, returns answers or routing actions to each channel, and sends performance feedback into the AI operations platform for retraining, flow updates, and release management.
As of November 2018, China Unicom served more than 400 million users (original: 截至2018年11月联通用户量超4亿户).
China Unicom used Baidu AI technology and the ABC smart customer service solution to optimize the basic service capabilities of the 10010 hotline and internet online channels (original: 借助百度人工智能技术、ABC 智能客服产品整体解决方案，中国联通优化了10010热线、互联网在线等不同渠道的基础服务能力).
It also provides an integrated platform for model training, process configuration, testing, release, analysis, and optimization, and runs on a container-based microservice architecture with elastic scaling (original: 提供可由企业自主模型训练、流程配置、测试发布、分析优化的一体化流程持续运营平台；基于容器的微服务架构，在高并发场景下可弹性伸缩).
The pain points included high channel-operating costs, isolated systems, lack of shared underlying capabilities, and old provincial hardware and software versions (original: 客服场景服务渠道及触点运营成本高，各渠道系统孤立建设、底层能力不共享、省份软硬件版本老旧等痛点).
The solution uses high-accuracy speech and semantic recognition, natural multi-turn dialogue, and a structured enterprise knowledge graph (original: 通过高准确的语音、语义识别技术、自然的多轮对话交互能力，依托结构化的企业知识图谱为用户提供智能贴心服务).
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 Jan 8, 2026 China Mobile Online (中移在线) China Mobile Online needed faster and deeper customer-intent insight than manual sampling and call listening could provide so it could support precision marketing, real-time service intervention, and low-threshold model… Clustering analysis timeliness is reduced to less than 1 hour from about 3 hours. Customer-Agent-Business Profiling Engine · Real-time Intelligent Insight Platform · Real-time Knowledge Push Engine Baidu AI Cloud Sample Jul 4, 2025 China Telecom China Telecom needed a differentiated anti-harassment service that could protect subscribers from nuisance calls, preserve privacy and service quality expectations, and scale across a nationwide telecom operation… Harassment-call interception rate increased to 98%. Baidu AI Cloud Intelligent Customer Service · Intent Analysis · Speech Recognition 
Cognizant Sample Archived · Oct 9, 2026 Cox Communications Cox Communications needed to modernize a high-volume chatbot and IVR estate so customers could get accurate, relevant answers across preferred channels without overloading live support teams. 17% improvement in first-contact resolution for chatbot Gemini Enterprise for Customer Experience (CX) · Google Cloud Contact Center as a Service (CCaaS) · Google Dialogflow CX 