nerdexam
Google

PROFESSIONAL-MACHINE-LEARNING-ENGINEER · Question #204

Your company stores a large number of audio files of phone calls made to your customer call center in an on-premises database. Each audio file is in wav format and is approximately 5 minutes long…

The correct answer is A. 1. Upload the audio files to Cloud Storage. To efficiently analyze 5-minute WAV audio files from on-premises using the Speech-to-Text API, upload them to Cloud Storage and use the asynchronous long-running recognition method.

Submitted by rohit_dlh· Apr 18, 2026Data processing and feature engineering

Question

Your company stores a large number of audio files of phone calls made to your customer call center in an on-premises database. Each audio file is in wav format and is approximately 5 minutes long. You need to analyze these audio files for customer sentiment. You plan to use the Speech-to-Text API You want to use the most efficient approach. What should you do?

Options

  • A
    1. Upload the audio files to Cloud Storage
  • B
    1. Upload the audio files to Cloud Storage.
  • C
    1. Iterate over your local files in Python
  • D
    1. Iterate over your local files in Python

How the community answered

(40 responses)
  • A
    70% (28)
  • B
    8% (3)
  • C
    18% (7)
  • D
    5% (2)

Why each option

To efficiently analyze 5-minute WAV audio files from on-premises using the Speech-to-Text API, upload them to Cloud Storage and use the asynchronous long-running recognition method.

A1. Upload the audio files to Cloud StorageCorrect

For audio files longer than 1 minute, the Google Cloud Speech-to-Text API requires asynchronous processing, necessitating uploading the files to Cloud Storage. The `long_running_recognize` method is specifically designed for transcribing longer audio files (up to 480 minutes) from Cloud Storage URIs, providing the most efficient approach for batch processing many 5-minute files.

B1. Upload the audio files to Cloud Storage.

The `stream_recognize` method is designed for real-time audio streams and typically has a 1-minute audio limit, making it unsuitable for pre-recorded 5-minute files.

C1. Iterate over your local files in Python

The `recognize` method is for synchronous requests and is limited to audio files 1 minute or shorter, making it unsuitable for 5-minute audio files.

D1. Iterate over your local files in Python

The `long_running_recognize` method requires a Google Cloud Storage URI as input for the audio file, not a local file path, so iterating over local files directly would not work.

Concept tested: Google Cloud Speech-to-Text asynchronous transcription

Source: https://cloud.google.com/speech-to-text/docs/sync-vs-async

Topics

#Speech-to-Text API#Cloud Storage#Data Ingestion#Audio Processing

Community Discussion

No community discussion yet for this question.

Sign up to join the discussion
Full PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice