Conversation Service
The Conversation Service is a single source of truth for all captured calls in the system, and makes use of various technologies and components like Cassandra and ClickHouse to power the different use cases across the U-suite products. When a call begins, the Conversation Service consults the Policy Manager to determine whether the call should be recorded. If recording is required, the Conversation Service assigns a universal Conversation ID and provides it to the Audio Connector. This ID is used to identify the call throughout the system, enabling consistent tracking, querying, and linking of related conversations.
The Conversation Service Upstream Consumer receives modernized metadata from the Audio Connector via GRPC. This metadata is mapped to expected metadata fields, which are uniform across all captured calls. The Conversation Service can then seamlessly store, query, and filter captured calls using this metadata. The Conversation Service Upstream Consumer also receives a transcription of the call audio from the ASR. This transcription is added to the call metadata using the Conversation ID.
When a user searches for a call in the UI, they can search across the metadata stored in the Conversation Service. The Conversation Service checks with the Policy Manager to confirm which calls the user is allowed to access (access privileges are determined using a combination of User Management Hierarchy/Roles and Access Control Policies, see Manage Users for Communication Recording Agent for details), and then provides filtered results to the user that match the users initial search and their access privileges.
Note
The available metadata for each call varies depending on the integration. See the relevant Technical Guidelines topic for details here.
Note
The above summarizes how call metadata is processed by the Conversation Service for a Communication Recording Agent system. The Conversation Service has additional features that are used by other applications.