A logistics company hosts live conference calls between English-speaking dispatchers and Spanish-speaking drivers and wants each spoken sentence translated into the other party's language and spoken back almost immediately during the live call, rather than produced afterward as a written transcript. Which AI capability addresses this requirement?
Select an answer to reveal the explanation.
Short Explanation
Focus on two details in the scenario: the input is spoken, not written, and the output has to arrive almost instantly, during the call itself, not afterward. That combination narrows things down fast. A capability that only handles already-typed text is out, because nobody typed anything here, the words were spoken live. A capability that just writes down what was said in the original language is out too, because writing it down in Spanish doesn't help an English-speaking dispatcher understand it. And simply figuring out which language someone is speaking doesn't translate a single word of it. What the dispatchers and drivers need is a capability built to take live speech in one language and hand back live speech or text in the other, fast enough to keep a conversation moving.
Full Explanation
The correct answer is B. Real-time speech translation is designed to take spoken audio as input and produce translated output, spoken or written, within the live flow of a conversation, which matches the requirement for near-immediate translation during an active call. Option A is incorrect because text translation operates on already-written text; it requires the spoken words to first be transcribed and would not, by itself, deliver a live spoken translation during the call. Option C is incorrect because speech-to-text transcription only converts audio into a written transcript in the same language it was spoken; it produces no translation at all, so a Spanish-speaking driver's words would remain in Spanish text. Option D is incorrect because language detection merely identifies which language is being used; it never converts or translates anything and would need to be paired with additional capabilities to be useful here. The specific combination of spoken input, translated output, and near-immediate delivery during a live call is exactly what real-time speech translation is built for.