14 July, 2025

Settings

We've added an option under "Settings > Workflows" to specify which tag events you want to consume on Intryc and make both visible on the UI and to the AutoQA. You can exclude or include all tag events or select the ones relevant to your QA criteria.

Data

We also have a new addition under help desk connection: Salesforce! You can integrate Salesforce with intryc and start evaluating your cases.

Evaluations

We have released an end-to-end flow for reviewing AutoQA evaluations. You have two options:

1. Go to the ticket that has the evaluation you want to review and select "Review AutoQA".

   

2. Go to "Evaluations > Completed by AutoQA" and select "Review AutoQA" next to the AutoQA evaluation you want to review. You can also filter evaluations based on whether they are reviewed or not.

   

In either case, the evaluation will open in edit mode and you can adjust and of the AutoQA output, including the score or "Skip" option, the comment, the root causes and the highlights. Once you're done, select "Submit review" and your changes will be applied. Reviewed evaluations will have the relevant tag next to the "AI Agent" with additional information about who reviewed it and when.

Reports

The best thing about reviewing AutoQA evaluations is that you can go to any of the AutoQA dashboard under Reports and select only the evaluations that have been reviewed. This allows you to review AutoQA performance and correction only across the evaluations your team has verified.

Sampling rules

We've made an update on sampling rules to consider by default only tickets which were created in the last 90 days. This aims to improve by up to x100 the performance of selecting random tickets for assignments and the number of available tickets in workloads.

Workloads

You now have more options and flexibility about the already evaluated tickets you can include in each workload.

Tickets

Manually created tickets are now shown instantly on the Tickets table!

Notifications

We've added a new notification type for admins when evaluations fail with 0% because of one ore more critical criteria.

We've also made minor improvements on all the notifications related to calibrations.

AutoQA

We've made important fixes on Knowledge Base retrieval for AutoQA.