Speech recognition turns spoken audio into written text. If an organization has hours of recorded interviews or customer calls, searchable transcripts make it easier to find a topic or quote without replaying every recording. Azure Speech to Text provides this capability. section 7.0 section 7.2
An interview archive
Transcribing author interviews lets readers search by keyword or topic and makes the recordings more accessible to hearing-impaired readers. Staff can also index quotes for marketing or use topics to trigger subscriber alerts. The transcript is useful well beyond the day the interview was recorded. section 7.2.1
Start by creating an Azure Speech services resource. The Azure CLI example selects the SpeechServices kind, the S0 pricing tier, a resource group, and a region. The resource can also support other voice-driven workflows, including Text to Speech and Speech Translation. section 7.2 section 7.6
C:/> az cognitiveservices account create --name banningBooksSpeech --resource-group BanningRG --kind SpeechServices --sku S0 --location westus --yes
Use this Azure CLI listing as a provisioning example; replace the resource name and group with your own. section 7.2
For a batch of recorded audio files:
- Retrieve the Speech resource key, then submit a transcription job using the batch API. Supply an accessible blob-container URL and a source locale such as en-US. section 7.2.1
- Request word-level timestamps if captions, jump-to-quote links, or synchronized highlighting need precise audio–text alignment. The example keeps results available for 48 hours. section 7.2.1
- Check the job status and follow the returned links to its output files. Submission is asynchronous, so a successful request does not mean the transcripts are ready yet. section 7.2.1
C:/> az cognitiveservices account keys list --name banningBooksSpeech
--resource-group BanningRG --query "key1" -o tsv
C:/> az rest --method post --url "https://banningbooksspeech.cognitiveservices.azure.com/
speechtotext/transcriptions:submit?api-version=2024-11-15"
--headers "Ocp-Apim-Subscription-Key=<your-key>"
"Content-Type=application/json"
--body '{"contentContainerUrl":
"https://<storage-account>.blob.core.windows.net/<container>?
<SAS-token>","locale":"en-US","displayName":
"BanningBooksInterviewTranscription","properties":
{"timeToLiveHours":48,"wordLevelTimestampsEnabled":true}}'
The listing shows key retrieval and a batch submission that processes audio from a blob container. Replace the placeholders before running it. section 7.2.1
C:/> az rest --method get --url "https://banningbooksspeech.cognitiveservices.azure.com/ speechtotext/transcriptions/<job-id>?api-version=2024-11-15" --headers "Ocp-Apim-Subscription-Key=<your-key>"
Use this GET request with the returned job ID to check transcription status and locate results. section 7.2.1
Speech recognition can also be the first step in Speech Translation: audio becomes text before translation produces captions or, optionally, dubbed audio. For a live event, streaming favors low-latency captions; for recorded material, batch processing leaves room for review before publication. section 7.2.3
