Deploying on-prem Speech services for Hybrid Meetings

Zoom AI On-Prem Speech Recognition services enable organizations to deploy Zoom AI-powered transcription and captioning capabilities within their own infrastructure for use with the Meetings Hybrid service. This solution is designed to support enterprises with strict data residency requirements, regulatory compliance needs, or security policies that mandate keeping sensitive meeting data on-premises or within their private cloud. These Speech services bring advanced automatic speech recognition (ASR) to on-premises or private cloud deployments via Zoom Node, allowing organizations to maintain control over their data while leveraging cutting-edge AI features.

Organizations can deploy these services across multiple Speech Zones, integrating them with existing Zoom Node infrastructure, including Meeting Zones and Recording Zones, to create a comprehensive hybrid meeting solution.

Learn more about managing captions as a host and viewing captions as a meeting participant.

Requirements for deploying AI On-Prem Speech Recognition services

• Business, Education, or Enterprise account
• Zoom Node subscription
• Account owner or admin privilege
• Deployed Node server based on the Zoom Hybrid Meeting system requirements

Limitations for deploying AI On-Prem Speech Recognition services

• On-prem Speech services only support the following languages:
  • English
  • Swedish
  • German
  • French

Deployment components for AI On-Prem Speech Recognition services 

When deploying the AI On-Prem Speech Recognition services, the following individual components will need to be deployed:

• Speech LTT Gateway: The external entry point that receives live captioning/transcription requests from Zoom Meetings Hybrid and routes them to a customer-controlled Speech Zone via the Mediator.
• Speech Mediator: An NGINX-based reverse-proxy that load-balances and routes traffic from the LTT Gateway to the appropriate ASR services based on language and availability.
• Secret Agent: A credential and token-management microservice that issues/validates ASR access tokens and serves as the trust anchor for all services in the Speech Zone.
• ASR Preprocessor: Prepares incoming audio streams for recognition by performing normalization, segmentation, and voice-activity detection before passing audio downstream.
• Language Identification: Detects the spoken language of an audio segment to determine which language-specific ASR pipeline should handle transcription (if available for that language).
• ASR Online: The real-time speech recognition engine that processes streaming meeting audio and produces live captions and transcripts.
• ASR Offline: A containerized batch-transcription service that processes recorded meeting audio after the sessions end.
• ASR Postprocessor: Refines raw ASR output by applying text cleanup, punctuation, and formatting before delivering final transcripts to downstream services.

How to add a Speech Zone

Before deploying the AI On-Prem Speech Recognition services, a Speech Zone must be configured and associated with a Meeting Zone and a Recording Zone. To add a new Speech Zone:

1. Sign in to the Zoom web portal as an admin.
2. In the top-right corner, click your profile picture or initials, then click Admin Center.
3. In the side menu, click Product Configuration .
4. Click Node, then click Modules.
5. At the top of the modules page, click the drop-down button , then click Meetings Hybrid.
6. On the Zones tab, click Speech.
7. Click + Add Speech Zone.
8. Under Zone name, enter the name of the Speech Zone.
   Note: This name is a unique identifier for the speech zone which will be shown in the Admin UI.
9. Under Zoom data center, select the data center you want to associate the Speech Zone with.
10. Under Meeting Zones, select the Meeting Zone(s) you want to associate the Speech Zone with.
11. Under Recording Zones, select the Recording Zone you want to associate the Speech Zone with.
12. Click Confirm. 

How to deploy AI On-Prem Speech Recognition services

Deploy the Speech LTT Gateway service

1. Sign in to the Zoom web portal as an admin.
2. In the top-right corner, click your profile picture or initials, then click Admin Center.
3. In the side menu, click Product Configuration .
4. Click Node, then click Modules.
5. At the top of the modules page, click the drop-down button , then click Meetings Hybrid.
6. On the Services tab, click Add Services.
7. Under Speech, click Speech LTT Gateway.
8. Under Install on a node, select the server you want to deploy the service on.
9. Under Speech Zone, select the zone you want to add the service to.
   Note: If you need to create a new Speech Zone, click New zone to create a new Speech Zone.
10. Click Confirm.

Deploy the Mediator service

1. Sign in to the Zoom web portal as an admin.
2. In the top-right corner, click your profile picture or initials, then click Admin Center.
3. In the side menu, click Product Configuration .
4. Click Node, then click Modules.
5. At the top of the modules page, click the drop-down button , then click Meetings Hybrid.
6. On the Services tab, click Add Services.
7. Under Speech, click Speech Mediator.
8. Under Install on a node, select the server you want to deploy the service on.
9. Under Speech Zone, select the zone you want to add the service to.
   Note: If you need to create a new Speech Zone, click New zone to create a new Speech Zone.
10. (Optional) Next to Add prefix, enter the certificate domain prefix.
11. Under Internal IP, enter the internal IP configured for the service.
12. (Optional) Under the SSL port, enter the port number the listening service should use.
    Note: The default port is 443.
13. Click Confirm.

Deploy the Secret Agent service

1. Sign in to the Zoom web portal as an admin.
2. In the top-right corner, click your profile picture or initials, then click Admin Center.
3. In the side menu, click Product Configuration .
4. Click Node, then click Modules.
5. At the top of the modules page, click the drop-down button , then click Meetings Hybrid.
6. On the Services tab, click Add Services.
7. Under Speech, click Secret Agent.
8. Under Install on a node, select the server you want to deploy the service on.
9. Under Speech Zone, select the zone you want to add the service to.
   Note: If you need to create a new Speech Zone, click New zone to create a new Speech Zone.
10. (Optional) Next to Add prefix, enter the certificate domain prefix.
11. Under Internal IP, enter the internal IP configured for the service.
12. (Optional) Under the SSL port, enter the port number the listening service should use.
    Note: The default port is 443.
13. Click Confirm.

Deploy the Preprocessor service

1. Sign in to the Zoom web portal as an admin.
2. In the top-right corner, click your profile picture or initials, then click Admin Center.
3. In the side menu, click Product Configuration .
4. Click Node, then click Modules.
5. At the top of the modules page, click the drop-down button , then click Meetings Hybrid.
6. On the Services tab, click Add Services.
7. Under Speech, click ASR Preprocessor.
8. Under Install on a node, select the server you want to deploy the service on.
9. Under Speech Zone, select the zone you want to add the service to.
   Note: If you need to create a new Speech Zone, click New zone to create a new Speech Zone.
10. (Optional) Next to Add prefix, enter the certificate domain prefix.
11. Under Internal IP, enter the internal IP configured for the service.
12. (Optional) Under the SSL port, enter the port number the listening service should use.
    Note: The default port is 443.
13. Click Confirm.

Deploy the Language Identification service

1. Sign in to the Zoom web portal as an admin.
2. In the top-right corner, click your profile picture or initials, then click Admin Center.
3. In the side menu, click Product Configuration .
4. Click Node, then click Modules.
5. At the top of the modules page, click the drop-down button , then click Meetings Hybrid.
6. On the Services tab, click Add Services.
7. Under Speech, click Language Identification.
8. Under Install on a node, select the server you want to deploy the service on.
9. Under Speech Zone, select the zone you want to add the service to.
   Note: If you need to create a new Speech Zone, click New zone to create a new Speech Zone.
10. (Optional) Next to Add prefix, enter the certificate domain prefix.
11. Under Internal IP, enter the internal IP configured for the service.
12. (Optional) Under the SSL port, enter the port number the listening service should use.
    Note: The default port is 443.
13. Click Confirm.

Deploy the ASR Online service

1. Sign in to the Zoom web portal as an admin.
2. In the top-right corner, click your profile picture or initials, then click Admin Center.
3. In the side menu, click Product Configuration .
4. Click Node, then click Modules.
5. At the top of the modules page, click the drop-down button , then click Meetings Hybrid.
6. On the Services tab, click Add Services.
7. Under Speech, click ASR Online.
8. Under Install on a node, select the server you want to deploy the service on.
9. Under Speech Zone, select the zone you want to add the service to.
   Note: If you need to create a new Speech Zone, click New zone to create a new Speech Zone.
10. Under Language, select the language to be utilized for transcription and captions.
11. (Optional) Next to Add prefix, enter the certificate domain prefix.
12. Under Internal IP, enter the internal IP configured for the service.
13. (Optional) Under the SSL port, enter the port number the listening service should use.
    Note: The default port is 443.
14. Click Confirm.

Deploy the ASR Offline service

1. Sign in to the Zoom web portal as an admin.
2. In the top-right corner, click your profile picture or initials, then click Admin Center.
3. In the side menu, click Product Configuration .
4. Click Node, then click Modules.
5. At the top of the modules page, click the drop-down button , then click Meetings Hybrid.
6. On the Services tab, click Add Services.
7. Under Speech, click ASR Offline.
8. Under Install on a node, select the server you want to deploy the service on.
9. Under Speech Zone, select the zone you want to add the service to.
   Note: If you need to create a new Speech Zone, click New zone to create a new Speech Zone.
10. (Optional) Next to Add prefix, enter the certificate domain prefix.
11. Under Internal IP, enter the internal IP configured for the service.
12. (Optional) Under the SSL port, enter the port number the listening service should use.
    Note: The default port is 443.
13. Click Confirm.

Deploy the ASR Postprocessor

1. Sign in to the Zoom web portal as an admin.
2. In the top-right corner, click your profile picture or initials, then click Admin Center.
3. In the side menu, click Product Configuration .
4. Click Node, then click Modules.
5. At the top of the modules page, click the drop-down button , then click Meetings Hybrid.
6. On the Services tab, click Add Services.
7. Under Speech, click ASR Postprocessor.
8. Under Install on a node, select the server you want to deploy the service on.
9. Under Speech Zone, select the zone you want to add the service to.
   Note: If you need to create a new Speech Zone, click New zone to create a new Speech Zone.
10. (Optional) Next to Add prefix, enter the certificate domain prefix.
11. Under Internal IP, enter the internal IP configured for the service.
12. (Optional) Under the SSL port, enter the port number the listening service should use.
    Note: The default port is 443.
13. Click Confirm.

Deployment of this service requires Zoom Professional Services.