Best Practices for Ticket Ingestion
The Ticket Ingestion process allows you to derive and analyze knowledge from your existing tickets.
There are few critical points to note about Ticket Learning. You only learn from tickets with these parameters:
the ticket has a status=CLOSED
the ticket must have non-empty title
the ticket must either have a Solution or at least one Comment
if both Solution and Comments are present, the learning algorithm only uses the Solution
The following information includes recommendations for Ticket Ingestion.
To ingest ticket data:
Choose the Settings > AiseraGPT command to open an application or bot that is integrated with a data source that contains the ticket data you want to ingest.
Edit the existing Data Source and make sure the Ticket Learning option is enabled, per the image below.

Select the Configuration tab and choose one of the following modes for data retrieval:

Date Range: each time the connector runs crawl all entities specified in its config that have been updated in the given range.
Incremental: in this mode the datasource will start from the time the last execution finished and crawl all the entities specified in its config that have been updated afterwards. If it is first time it will retrieve all entities specified in its config it will find in the external system without date constraints.
Hit the Play button of the Data Source to kick-off ticket ingestion and wait until it finishes.
Last updated