Inbound Email Action – parsing HR emails (hidden characters, cut values, location match)

[Community Alums]

Use case
Our HR system (SuccessFactors) sends onboarding and termination emails. An Inbound Email Action reads the email, creates a RITM, and a flow creates regional tasks based on location.

Problems
1. Labels like "Employment Type" were not found.
2. Values with a colon were cut. Example: H:\Accounting became H.
3. Location was empty or matched a retired site.
4. The flow did not start its regional branch.

Root causes
1. The emails contain hidden zero-width spaces and non-breaking spaces, so label matching failed.
2. The script split each line on every colon.
3. Site names in the email did not exactly match cmn_location names (accents, short names). The active flag on cmn_location was a custom field, so the normal active filter did not work.
4. The location variable was a Choice field, not a Reference to cmn_location, so the flow could not dot-walk to Region.

Fix
1. Clean the email body first: remove zero-width characters and replace non-breaking spaces with normal spaces.
2. Split each line on the first colon only.
3. Change the location variable to Reference, pointing to cmn_location.
4. Match location step by step, so it is never empty:
- exact name match on active sites
- alias list match
- closest active match by shared words
- log unmatched values for review
5. In the flow, branch on location.region.

Tips
- Build the alias list from old RITM history.
- Test with accented site names.
- Check which field means "active" on your cmn_location table.

Hope this helps. Happy to share more details if needed.