Customer Service Automation
vivo wanted its BlueLM-powered phone assistant to answer general questions while also retrieving current and authoritative information without forcing users to open browsers or other apps.
Reported impact
vivo upgraded its assistant from a general-purpose model interface to a real-time service surface that saves users time, improves accuracy and strengthens the brand's AI-phone positioning.
vivo integrated Baidu AI Cloud's Qianfan platforms into the BlueLM and BlueLM Assistant stack so its phone assistant could mix model reasoning with fresh web knowledge. The core architectural change was the introduction of a retrieval-augmented generation path connected to Baidu Search. When users ask about weather, stocks, exchange rates, policy changes, travel or current affairs, BlueLM Assistant can invoke Baidu Search-sourced retrieval components rather than relying only on a static device model. The result is a hybrid design in which vivo keeps its own BlueLM device-side experience and assistant persona, while Baidu provides model services and the RAG pathway that grounds answers on recent information with source traceability. The same collaboration also explores a broader phone-agent ecosystem, positioning the assistant as a front door for more specialized mobile intelligence rather than only a chat interface.
Phone-assistant questions are routed through BlueLM Assistant; when a topic needs timely information, RAG with Baidu Search retrieves fresh results, which the model stack summarizes and returns inside the phone assistant without forcing a browser handoff.
The challenge was enabling BlueLM Assistant to fetch the latest internet information while preserving general Q&A abilities (original: 首先面临的主要挑战是如何让“蓝心小V”在保持通用问答的基础上，实现实时获取互联网最新资讯和权威内容的能力).
The solution improves retrieval for weather, stocks, exchange rates, policy, travel and current events without requiring users to open a browser (original: 针对天气、股票、汇率、政策、出行、时事新闻等领域迅速、准确地为用户提供所需信息，无需跳转浏览器).
vivo used Qianfan together with the RAG with Baidu Search components to connect to Baidu Search-sourced real-time content (original: 通过RAG with Baidu Search系列组件接入与百度搜索同源的实时联网内容).
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 Jun 20, 2025 OPPO OPPO needed to bring trustworthy real-time information, personalization, multimodal understanding and developer tooling into its AI-phone strategy while still protecting user data and keeping assistant experiences… Baidu Search-Sourced Retrieval · ERNIE Large Model · Qianfan AppBuilder Baidu AI Cloud Sample Jun 16, 2025 Shifeng Culture Development Co., Ltd. Shifeng Culture wanted to turn traditional children's toys into interactive AI products, but needed help with large-model integration, low-latency interaction, security and scalable product delivery. Cloud Security Services · Intelligent Interaction Platform · Qianfan AppBuilder Baidu AI Cloud Sample May 15, 2025 Samsung Electronics China Samsung needed to bring reliable generative-AI capabilities into its China smartphone lineup, covering typing, office work, voice assistance and call handling while reducing hallucinations and unstable output formatting. Voice-assistant comprehension improved 50%. Bixby · Galaxy AI · Qianfan ModelBuilder 