引導式測驗 · 中等 等級

銷售聊天機器人建築服務測驗

圍繞真實任務、經批准的知識、人工升級、可衡量的測試、資料保護和持續支援來確定聊天機器人服務的範圍。

問題 1 的 8

What should a chatbot service provider clarify before quoting a build?

預覽每個問題8 問題

當您參加測驗時,就會出現答案和解釋。

  1. What should a chatbot service provider clarify before quoting a build?
    • A guarantee that the bot answers every question
    • The model name alone
    • User task, approved sources, integrations and failure handling
    • A promise to remove all human support
  2. Why include an escalation path in a chatbot?
    • It guarantees hallucinations never occur
    • Some questions are uncertain, unsupported or consequential
    • It removes the need to test
    • It automatically updates the knowledge base
  3. What should a pilot measure?
    • Count messages handled without checking whether requests were resolved
    • Measure a hypothetical time saving based on a vendor benchmark
    • Measure only deployment time, excluding recurring failures
    • Task completion, answer quality, escalation, and unresolved requests
  4. Why version the chatbot's source documents?
    • To trace which content supported an answer and manage updates
    • To guarantee every answer is correct
    • To replace access controls
    • To prevent users from asking questions
  5. Which ongoing service responsibility should be scoped?
    • Only choose a chat-window color and brand style
    • Only set the initial prompt once
    • Who updates content, monitors failures, and provides support
    • Only include the model’s one-time training fee
  6. How should a chatbot handle a consequential action such as changing an account?
    • Execute any generated instruction automatically
    • Use explicit permissions, validation and confirmation
    • Store credentials in the user prompt
    • Skip audit records
  7. How should a service provider communicate expected business outcomes?
    • Guarantee savings before testing
    • Claim full autonomy from a prototype
    • Use provider benchmark numbers as client results
    • Report measured results and disclose limitations
  8. When should an organization reconsider using a chatbot for a task?
    • When errors could cause serious harm and qualified review is absent
    • When the bot can retrieve approved documents
    • When support agents can take handoffs
    • When the system has a test set