AI-900 Practice Question: Describe features of Natural Language Processing workloads on Azure

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Real-time transcription of live audio streams for immediate use

  Why it's wrong here

  Batch transcription is designed for pre-recorded audio files submitted in bulk, processed asynchronously, with results retrieved later via status polling. Real-time transcription streams live audio through a low-latency API (e.g., WebSocket) to produce immediate captions or subtitles; this is a distinct feature and not what batch transcription does. Batch mode lacks the real-time return of word-by-word results.

• ✓

  Asynchronous processing of large volumes of audio files for cost-efficient transcription at scale

  Why this is correct

  Batch transcription is correct because it is specifically built for asynchronous, high-volume transcription: you upload a container of audio files (via a shared access signature URI), the service processes them asynchronously, and you later poll for and retrieve the stored transcripts. This scale-oriented design is cost-efficient and ideal for call-center archives, media catalogs, or meeting recordings that need post-hoc analysis. It is the dedicated bulk operation in Azure Speech-to-Text.

• ✗

  Synchronizing speech transcription across multiple languages simultaneously

  Why it's wrong here

  Batch transcription transcribes audio per-file using a specified source language locale; it does not synchronously produce multi-language transcripts or translations. Multi-language simultaneous transcription or translation is handled by separate Azure capabilities like speech translation or multilingual conversation transcription, which address language coverage rather than bulk asynchronous volume. Selecting this option mistakenly equates 'batch' with 'multilingual,' but batch refers to processing scale, not language diversity.

• ✗

  Training a custom speech recognition model on audio samples

  Why it's wrong here

  Batch transcription is an asynchronous service that applies the Speech service's existing recognition models to pre-recorded audio files; it does not train or adapt models. Custom speech model training is a separate capability where you upload labeled audio or text to build a personalized acoustic/language model, then deploy that model for recognition. Choosing this option confuses model creation with bulk batch processing.

Go deeper

Related to this question