To ensure the proper operation of AI features such as speech transcription, intelligent meeting summaries, and real-time subtitles, it is mandatory to configure both a public cloudobject storage service and an AI model service.
When utilizing speech-related AI services, the system uploads local audio files to the configured public cloud object storage, generating publicly accessible links. The AI model accesses these links to retrieve and process the audio content. Once processing is complete and results are returned, the system immediately deletes the audio files from the cloud storage, ensuring data security and optimizing storage usage.
Currently, Deepgram is the only AI model service that supports direct local audio file uploads, eliminating the need for public cloud object storage configuration. If you exclusively use Deepgram, you may skip the object storage setup.
Below are the supported object storage services along with links to their official configuration documentation:
Please select the appropriate object storage service based on your requirements and follow the official documentation to complete the configuration.
Below are the supported AI model services along with links to their official API Key acquisition pages:
Please select the appropriate AI model service based on your requirements and follow the official documentation to acquire and configure the API Key.
After completing the above configurations, your system's AI features should function correctly. If you have any questions or require further assistance, please contact our technical support team.