Amplify Enrich FAQ

Got questions about Amplify Enrich? You're in the right place! This guide contains answers to common questions about Amplify Enrich.

See Amplify Enrich Troubleshooting for help solving common issues with the Enrich step.

Enrich Automation Step

How does the Enrich Step work?

Here's the basic flow:

  1. Add the Enrich step to a candidate-, submission-, or job-based automation.

  2. Pick your prompt - choose from our default prompts or create your own custom ones

  3. Choose where to save the AI-generated content:

    • Directly to an ATS field, or

    • As a note on the candidate or job record.

Once a record passes through the Enrich step, your prompt runs automatically and the content gets saved exactly where you specified.

Which automations can I use the Enrich Step in?

You can use the Enrich step is Candidate and Job based automations, as well as submission-based automations with a Type of ‘Candidate’. The Enrich step is available for both standard and date-based automations.

How do I create my own prompts for the Enrich step?

Admins can use the Prompt Studio in Bullhorn ATS to build custom prompts. Once they’re created, they'll automatically show up as options in the Enrich step for the relevant entity.

How can I tell when a record has been enriched?

Just check the Activity Feed on the record in Automation! Here's what the statuses mean:

  • "Waiting": The record has been processed and is waiting for Amplify to complete the request.

  • "Completed": Amplify has finished and the content has been generated.

Why aren't all the default prompts in Bullhorn ATS available for use in Automation?

Bullhorn Automation currently supports prompts that are:

  • Entity-specific (designed for candidates, jobs, etc.)

  • Scalable across multiple records rather than tailored to individual records

This approach keeps your automated actions relevant, efficient, and effective at scale. Prompts that work best for one-off, individual record scenarios in the ATS aren't quite the right fit for Automation workflows.

What fields can the Enrich step write to?

You can choose any text field in the ATS. The AI-generated content comes back as a text string, so it needs to fit within the character limits and format requirements of your selected field.

 

  • Existing data will be overwritten - If your selected field already has content, the AI-generated text will replace it

  • Format matters - Content won't save if the field format is incompatible (for example, you can't save a job title to a date field)

Why can't the Enrich step write back a list of skills or categories to the standard ATS fields?

This comes down to how Bullhorn handles structured data. The standard skills and categories fields require each item to have a specific ID for validation—they're not simple text fields.

Right now, Automation delivers AI-generated content as plain text (like "JavaScript, Python, Project Management"), which doesn't match the structured format the ATS needs. We're working with plain text output, but those fields need specially formatted data with unique IDs.

How do I stop the Published Description field from automatically copying the Job Description?

By default, the Published Description field automatically fills with whatever is in the Job Description. If you're using an Enrich step to generate enhanced published descriptions, this can make it harder to see which jobs have not yet been enriched.

To turn off this behavior:

  1. Log in to your ATS as an Admin user.

  2. Navigate to Menu > Admin > System Settings.

  3. Use the search bar to find autoPopulatePublicDescription.

  4. Set the value to FALSE.

  5. Save your changes.

Once updated, the Published Description field will no longer autopopulate, allowing your Enrich step to generate the enhanced content instead.

Enrich From File Automation Step

What happens if no resume data or matching file is found?

The candidate is removed from the automation step.

How does the system determine which file is "most recent"?

The system uses the dateAdded field. If multiple files match the selected file types, whichever was added most recently is used.

Can I update individual fields within a work history or education entry?

No. Work history and education are processed as complete sections. The system cannot update individual fields within these sections. Either the entire section is updated, or it is not touched.

Why does the system replace data instead of adding to it?

The system replaces data because a resume is meant to reflect a candidate’s current qualifications. Tools like Search and Match rely on up-to-date skills and experience to determine job fit, and mixing older resume details with newer ones can reduce match quality. The candidate may no longer consider older information relevant to their current experience level or career focus.

What if I want to keep some existing data but update other fields?

Use the "Only update when empty" setting for this. Standard fields, like phone or email, will only change if they’re empty. However, for skills, work history, and education, the system treats each section as a whole, so if any information already exists, that section won’t be updated.

Why did the Enrich From File step update the wrong field?

Enrich From File always populates the default Bullhorn field shown in brackets in this table. If your workflow uses a different or renamed field (including custom fields), that field will not be updated.

To resolve this, check the table and ensure you are using the Bullhorn default fields supported by Enrich From File. Custom mappings are not supported.

I can’t see the information populated by the Enrich From File step. Where did it go?

If the enriched data isn’t visible on the candidate record, check the following:

Once these fields are visible and mapped properly, the enriched data will appear as expected.

Does Enrich From File support custom fields?

No. Enrich From File only writes to the default Bullhorn fields shown in brackets in this table. Custom fields and custom mappings are not supported.

Enrich Records from Notes in ATS

Which fields can Amplify update from notes?

An admin can configure up to 20 fields per entity (Candidate, Contact, or Job). Amplify only suggests updates for the fields selected in Amplify Admin.

Does Amplify automatically apply field updates?

No. You choose which updates to apply. You can also edit any suggested value before saving.

Can low-confidence suggestions still be applied?

Yes. Low-confidence suggestions are unchecked by default, but you can select them and adjust values as needed before saving.

Can admins control which fields are auto-selected when Amplify makes suggestions?

Yes! Admins can use the Auto-check toggle in Amplify Admin > Entity Settings to set the default selection behaviour for each field, so you can choose whether a high-confidence update to a particular field comes pre-selected or whether users need to check it manually before applying an update.

Can I enable Note Enrichment for only certain users or departments?

No. Note Enrichment is enabled at the corporation level and applies to all users once turned on.

Does editing an existing note trigger suggestions?

Yes. Editing any existing note on a candidate, contact, or job can trigger new or updated suggestions based on the revised text.

Do third-party note takers trigger field updates?

No. Notes created automatically by third-party note takers don’t trigger updates. Amplify only analyzes and suggests updates after a user edits the note, so there’s no risk of records changing based solely on a notetaker’s output.

Best practice is for users to review the autogenerated note after the call, make any necessary edits, verify the suggested updates, and then save.

Enrich from Note Automation Step

Which note action types should I select?

Select the action types that are most likely to contain structured candidate information, such as intake call notes or interview summaries. The step only processes notes that match the action types you configure, so choosing relevant types helps ensure the right notes are used.

What fields can the Enrich from Note step update?

The fields available for extraction are determined by your Enrich Fields configuration in Amplify Admin. An admin can configure up to 20 fields per entity, including custom fields. See Configuring Enrich Fields for setup steps.

What happens if no matching note is found?

If no note is found that matches the selected action type(s) with the candidate in the About field, the candidate is skipped for that step.

What is the difference between the confidence threshold options?

Amplify assigns a confidence level to each piece of data it extracts from a note. The When to Update setting controls the minimum confidence level required for a field to be updated:

  • High applies the fewest updates but with the greatest accuracy.

  • Medium balances coverage and accuracy.

  • Low applies the most updates but increases the risk of inaccurate data being written to candidate fields.

If you're not sure which to use, start with High confidence or higher and adjust from there.

How is the Enrich from Note step different from enriching a candidate from a note manually in the ATS?

When you add or edit a note manually in the ATS, Amplify suggests field updates and you choose which ones to apply before saving. The Enrich From Note automation step applies the same extraction automatically, based on the confidence threshold you configure. There is no manual review step, and updates are applied directly to the candidate record. This makes it well suited for processing notes in bulk or applying enrichment to historic records.

How can I tell what the Enrich From Note step updated on a candidate record?

To see what was changed, check the Edit History on the candidate's Activity tab. Updates applied by the Enrich From Note step are attributed to the Automation API User.

Why did the Enrich from Note step not update a record as expected?

Check the following:

  • The fields you expect to update are included in your Enrich Fields configuration in Amplify Admin.

  • The note matches one of the selected action types and the candidate is listed in the About field (not Additional References).

  • The extracted value met or exceeded the configured confidence threshold.