The GET /Conversations/Classifers/ end point returns the catalog of classifiers available for your organization. These are the AI models, LLM evaluations, and metadata scorers that have been applied to conversations.
Use this endpoint as a discovery step before querying conversation results. Because classifier IDs are org-specific, call GET /Conversations/Classifiers/ first to map classifier names to their Id values, then pass those IDs when filtering or interpreting data from downstream endpoints such as /Conversations/Insights/ or /ModelResults/.
The GET /Conversations/Insights/ end point returns AI model predictions, skill detections, and topic classifications for processed conversations within a given date range. For each conversation, the response includes the full set of classifier results (each with the classifier name, type, category, and result value) alongside the conversation identifier and dates needed to cross-reference with other endpoints.
Results can be filtered by specific classifiers, classifier categories or types, and organizational hierarchy, making this the primary endpoint for exporting conversation analysis outcomes in bulk.
Each Insights response includes a
ConversationId. This is the same conversation identifier returned asIdfrom the Metadata endpoints, and can be passed directly to the Transcriptions endpoint to retrieve the full transcript for that conversation.
