Understanding Import Types
In Voxjar, imports are used to bring data into your account in bulk. There are two types of imports:
Ad hoc Imports: These are for immediate processing of data.
Auto Imports: These run automatically at set intervals or in realtime.
Creating an Import
Navigate to the Imports page.
Click "Create a new Import" from the top right.
Choose your data source. Options include:
Selecting data from existing uploads
Choosing data from integrations
Uploading URLs or files directly
Set filters and rules to determine which interactions and how many interactions to import.
If you are importing from an Interaction you will have additional options to determine if your import should run automatically (on a schedule or in realtime).
Select your evaluators. You can choose between:
AI evaluation
Human evaluators
If selecting human evaluators, you can define distribution rules for how the interactions are assigned.
Managing Imports
Once your imports are set up, you can manage them in several ways:
View import details: See information about each import, including its status and scheduled run times.
Monitor progress: Check the status of currently running imports.
Export results: Download the data from completed imports.
Archive imports: Move completed or unnecessary imports to an archive for cleaner organization.
Edit imports: Make changes to existing imports as needed.
For Ad hoc imports, you can view the details of the queue, but since they've already run, you can't edit them.
For Auto Imports, you can edit all aspects of the import, including the data source, schedule, and evaluator settings.
Import Processing and Limitations
It's important to note that there is a limit to how much data can be processed in one queue run:
This can vary based on the data source but the typical maximum number of interactions that can be processed in a single run is 10,000.
If you need to process more than 10,000 interactions, you can create multiple imports.
When an import has completed, it means that all items have been queued up for processing. It doesn't necessarily mean that all evaluations have been completed. You can check the status of individual evaluations in the queue details.