Uploaded October 2026 | Updated November 2026, 2 days ago
Add OCI AI Speech Live Transcription capabilities to your APEX applications with practical reference architecture for browser audio capture, secure WebSocket streaming, and real-time transcript results. Learn how an application obtains a session token through a Speech REST API, streams microphone audio to the service, displays partial and final text, and saves completed transcript segments.
Chapters:
0:00 Live speech-to-text reference architecture
1:10 Recording session lifecycle
3:00 Page, server, and browser responsibilities
4:30 Configure recognition settings
6:15 Create session tokens and authenticate WebSockets
7:30 Demo the live transcription application
9:45 Review scripts and transcript persistence
11:30 Production considerations and wrap-up
This webinar explains how APEX can coordinate a live transcription session while keeping private signing credentials on the server. The browser captures microphone audio in a secure context, opens the WebSocket connection with a session token, renders partial results as a preview, appends finalized segments to the transcript, and sends completed segments for database storage.
Learn the key API settings that shape recognition: audio encoding, model type and domain, language, punctuation, partial and final silence thresholds, partial-result stability, customizations, and service acknowledgements. The video also covers token attributes, response metadata, client-side controls, server-side REST calls, and a practical pattern for saving final transcript text.
Key links:
GitHub repository and application scripts: github.com/uzer15/Oracle/blob/main/APEX/AI%20Speech%20Live%20Transcription.md
AI Speech documentation: docs.oracle.com/en-us/iaas/Content/speech/using/using-live-transcribe.htm
Speech API: docs.oracle.com/en-us/iaas/tools/python/latest/api/ai_speech/models/oci.ai_speech.models.RealtimeSessionToken.html
Explore more Code Innovate videos:
youtube.com/playlist?list=PLPIzp-E1msrZMCfSHbKgLK3KWsNM9JB9a
#codeinnovate #SpeechToText #LowCode #WebSockets #AI
Add OCI AI Speech Live Transcription capabilities to your APEX applications with practical reference architecture for browser audio capture, secure WebSocket streaming, and real-time transcript results. Learn how an application obtains a session token through a Speech REST API, streams microphone audio to the service, displays partial and final text, and saves completed transcript segments.
Chapters:
0:00 Live speech-to-text reference architecture
1:10 Recording session lifecycle
3:00 Page, server, and browser responsibilities
4:30 Configure recognition settings
6:15 Create session tokens and authenticate WebSockets
7:30 Demo the live transcription application
9:45 Review scripts and transcript persistence
11:30 Production considerations and wrap-up
This webinar explains how APEX can coordinate a live transcription session while keeping private signing credentials on the server. The browser captures microphone audio in a secure context, opens the WebSocket connection with a session token, renders partial results as a preview, appends finalized segments to the transcript, and sends completed segments for database storage.
Learn the key API settings that shape recognition: audio encoding, model type and domain, language, punctuation, partial and final silence thresholds, partial-result stability, customizations, and service acknowledgements. The video also covers token attributes, response metadata, client-side controls, server-side REST calls, and a practical pattern for saving final transcript text.
Key links:
GitHub repository and application scripts: github.com/uzer15/Oracle/blob/main/APEX/AI%20Speech%20Live%20Transcription.md
AI Speech documentation: docs.oracle.com/en-us/iaas/Content/speech/using/using-live-transcribe.htm
Speech API: docs.oracle.com/en-us/iaas/tools/python/latest/api/ai_speech/models/oci.ai_speech.models.RealtimeSessionToken.html
Explore more Code Innovate videos:
youtube.com/playlist?list=PLPIzp-E1msrZMCfSHbKgLK3KWsNM9JB9a
#codeinnovate #SpeechToText #LowCode #WebSockets #AI










