Predictive Intelligence - Flow Designer Actions Not Working (San Diego)
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
12-16-2022 06:10 AM
Hi,
Having followed the steps outlined in this youtube video which covers the steps required to create a PI Classification Solution for automated assignment of assignment group and integrating this into the process using Flow Designer PI Actions. https://www.youtube.com/watch?v=-s5ECLKkKlM
After following these steps exactly within our instance, we are finding that the flow is not executing correctly.
The issue we are finding is that even when the Confidence to Predict value (string) is False (Confidence = below comparison threshold), the flow is still going down the path it would if it was True. See below:
For each prediction result:
We are performing a PI Confidence Check, taking the Confidence value, and using the Comparison Threshold of 40:
The If condition looks at the Confidence To Predict value (string, but result is either True or False) and looking for the output to be True:
If this is true, then it will add a work note, update the assignment group and update a field on the Incident record:
Then we have an Else If to take the opposite position, Confidence To Predict is False, then it adds a work note.
When we run this flow through creating a new Incident, this is what we get, based on the short description there is a confidence value of 78.2:
Which results in Confidence To Predict = True:
But even though the runtime value comes back as True=True, the Flow Logic comes back as Evaluated - False, and therefore the actions are not performed:
It also doesn't trigger the Else If:
I have seen a number of references in other community articles to a bug within ServiceNow with regards to If statements being used for string values.
Is this something that other people have come across and if so have you managed to find a way around it? Right now this is preventing us from implementing PI.
Thanks in advance.