Martin Sachs

Hi Tetiana Gron  - in May, you left a comment that the generative answer comes from the top 3 articles.  If this is the case, how is the linked article chosen from the 3?  Our testing is showing us that we are getting correct answers but the linked article isn't where the answer came from.  It is returned in the top 3 and has a few keywords but the response itself is clearly from one of the other top 3 articles and not the one AI displays.   We are seeing this same behavior in the Agent Generative search as well.  

We'd love to understand this more as we plan for changes to our content structure.  

It appears that an internal comment deactivates suggested replies.  This is problematic because of years of automatons that include private comments to alert agents to conditions when tickets are created.   Private comments are also added if the ticket needs to be escalated to another group.   It makes perfect sense to disable suggested replies once there is a public reply to the End User but I don't see values in disabling this based on Private comments.

If this was intentional, I'd love to understand the rationale that I am not considering.

I was able to do this - for me to add most of the post steps, I go to the text version where it shows all the transitions and create is not there.  If I switch to the diagram option, I see a workflow step for create, can click there and add the post function.   I didn't test it all the way through but I'm pretty sure this worked for me years ago.   Let me know!

 

We are in need of this as well.  It becomes very challenging to meet the reporting needs without this.  I believe this level of granularity was available when Good Data was the BI solution. 

I'm reading this last comment above and it explains why I'm stuck.  Does this mean that we can only find articles with no views for all time vs recently?  I know the engagement filters go back to 2020 but we are looking for low or no view articles only since 2022.  Am I correct that we can't achieve this in Explore?  I'm thinking getting views with filters and then a standalone report with all published article IDs allowing a Vlookup and assuming null views for the delta is the workaround.  Does that sound logical?

Sarah Stocks  Happy to connect assuming that was directed to me.  I also tried to private message you but couldn't.  Maybe our Zendesk CSMs can put us in touch?

I'm not sure an automation would work here.  Last I checked Zendesk doesn't support placeholders or formulas in triggers or automatons so you'd effectively have to have one for each org.  If Org is A then ticket org A etc etc.   That is not possible in my world thus the webhook route where the JSON payload can reference an Org placeholder. 

Also, once you do get the value at the ticket level fields, you can include it in mappings to Jira assuming your field on that end is a matching text type.

Does that make sense?  If I'm wrong about not being able to use triggers or automatons, Id love to learn because this has been a challenge of mine with both Jira and Salesforce integrations. 

HI Sarah Stocks,

From my experience, the field sync only supports ticket level fields, not Org fields.  The org field can be mapped in the iFrame portion and I have Org Visible in Jira but not in a native Jira field so you can't filter or build lists and reports on it - just view it.

I solved this by using a webhook from Zendesk to Zendesk that puts the organization name as a text custom field on all tickets when created.  This then allows that to sync to a text field in Jira.  The complication here is it consumes API threshold limits so as you add more and more of this, you have to monitor your API consumption and rate limits.  We combine multiple needs into one webhook to minimize this.

Hope that helps - happy to share some examples of how we do this.

This is indeed very exciting.  We are heavy users of the Jira integration and use the post function on the jira side to reopen Zendesk tickets when they transition to certain statuses in Jira.  I assume this uses the API.  Will this integration continue to work as expected once activated since I don't imagine the UI in Jira will NOT update to reflect categories vs status?

Similar question for the Salesforce Integration.  ZD Tickets are sync'd to SFDC cases - will the new status vs categories be fields we can sync in that integration?

Dave Jensen - avail in the native integration but only works if there is one Jira tied to the Zendesk.  You know where to find me if you want some help.  Hope you're well