A company is gathering audio, video, and text data in various languages. The company needs to use a large language model (LLM) to summarize the gathered data that is in Spanish. Which solution will me

Question

Options

• ATrain and deploy a model in Amazon SageMaker to convert the data into English text. Train and
• BUse Amazon Transcribe and Amazon Translate to convert the data into English text. Use Amazon
• CUse Amazon Rekognition and Amazon Translate to convert the data into English text. Use
• DUse Amazon Comprehend and Amazon Translate to convert the data into English text. Use

• A
  9% (4)
• B
  72% (34)
• C
  15% (7)
• D
  4% (2)

Explanation

Option B is correct because it pairs the right AWS services for each data type: Amazon Transcribe converts audio and video speech into text (handling the non-text modalities), while Amazon Translate then converts that Spanish text into English - all without any model training, making it the fastest path to a working solution.

Option A fails the "least amount of time" requirement because training and deploying custom SageMaker models is time-intensive compared to using pre-built managed services.

Option C is wrong because Amazon Rekognition analyzes images and video for visual content (objects, faces, labels) - it cannot transcribe spoken audio, so audio data would be unhandled.

Option D is wrong because Amazon Comprehend performs NLP analysis (sentiment, entities, key phrases) on already-existing text - it cannot ingest audio or video and convert them to text.

Memory tip: Match the modality to the service - "Transcribe = Transform speech to text, Translate = Transform language, Rekognition = Rekognize visual content, Comprehend = Comprehend meaning from text." When you see audio + video + text together, Transcribe is almost always the right first step.

No community discussion yet for this question.