> For the complete documentation index, see [llms.txt](https://docs.maiagent.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.maiagent.ai/maiagent-user-guide/maiagent-user-guide-en/application/voicecs/call-qa.md). # Voice Call Quality Inspection This chapter introduces how to build an intelligent voice call quality inspection AI assistant that combines Automatic Speech Recognition (ASR) technology to transcribe call content between agents and customers into text, and then automatically reviews service quality. This helps supervisors quickly understand customer service performance and avoid missing potential service flaws or compliance risks. The AI assistant needs to have the following core functions: 1. **Service Standard Compliance Check**: Based on the company's established customer service SOP, evaluate item by item whether agents have completed required scripts, such as self-introduction, customer identity verification, and closing courtesy. 2. **Sentiment and Attitude Analysis**: Analyze the agent's tone and word choice to identify situations such as impatience, argumentation, or unprofessional behavior, while also tracking changes in customer sentiment. 3. **Compliance and Forbidden Phrase Detection**: Automatically check whether the call content contains promissory language, sensitive terms, or scripts that violate regulations, reducing legal compliance and complaint risks. 4. **Structured Inspection Output**: Output inspection results in structured fields, making it easy to import into back-end systems or BI reports, helping supervisors aggregate and track service quality. With fully automated AI quality inspection, the burden of manual sampling is significantly reduced. Where previously only a small number of calls could be sampled, now every call can be reviewed in real time, helping enterprises identify service gaps and make improvements at the earliest opportunity. ### Role Instructions ```markdown # Role You are a call quality inspection assistant for the enterprise customer service center. The input is the transcript of a call between an agent and a customer. # Inspection Items 1. Opening: Whether the agent provided a complete self-introduction (company name, agent name/ID). 2. Identity Verification: Whether the agent verified the caller's identity according to standards. 3. Service Attitude: Whether the agent's tone is friendly, with no impatience or argumentation. 4. Forbidden Phrase Detection: Whether any promissory, aggressive, or non-compliant language appeared. 5. Closing Courtesy: Whether the agent confirmed that the customer's issue had been resolved and offered a proper farewell. # Output Constraints 1. Each inspection item can only be marked as "Pass", "Fail", or "Not Applicable". 2. If failed, please supplement the original sentence and reason in the "Specific Notes" field. 3. For fields that cannot be determined, please enter "None". # Output Format Please output directly without tags. ``` ### Voice Conversation Text ``` Agent: Hello, this is MaiAgent Customer Service Center. I'm Mei, agent ID A123. How may I help you? Customer: Hi, the amount on last month's bill is incorrect. Why was I charged an extra five hundred dollars? Agent: I sincerely apologize for the inconvenience. To protect your rights, may I first verify your identity? Please provide your member ID or registered mobile phone number. Customer: My phone number is 0912-345-678. Agent: Thank you, I have located your account information. I can see that there is indeed a five-hundred-dollar value-added service charge on last month's bill, which was generated by the "Family Protection Plan" you activated on our website on November 5th. Customer: I never activated that plan! Are you charging me arbitrarily? Agent: I understand your concern. I will immediately help you retrieve the activation record and report back. A specialist will contact you within 24 hours with a resolution plan. May we reach you at this mobile number? Customer: Yes, thank you for your help. Agent: A case has been created for you. Is there anything else you need assistance with? Customer: No, thank you. Agent: Thank you for calling MaiAgent Customer Service Center. Have a great day, goodbye. ``` ### Quality Inspection Result ```markdown Opening: Pass Identity Verification: Pass Service Attitude: Pass Forbidden Phrase Detection: Pass Closing Courtesy: Pass Specific Notes: The agent provided a complete company name and agent ID, and followed standards by asking for the customer's mobile number for identity verification. When facing the customer's questioning, the agent remained calm, avoided argumentation or promissory language, and proactively created a follow-up case. Overall Recommendation: This call complies with service standards. It is suggested that when responding to customer concerns, the agent could further explain the specific process for retrieving records to strengthen customer trust. ``` --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.maiagent.ai/maiagent-user-guide/maiagent-user-guide-en/application/voicecs/call-qa.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.