Skip to main content

How is interview feedback summarized in PI Hire?

Learn how we use artificial intelligence (AI) to summarize interviewer feedback in PI Hire.

Updated over 3 weeks ago

This article applies to all Hire users in the PI2 platform.

The Interviewer Feedback page in Hire includes the ability to generate an AI-powered summary of your team's interview notes.

In this article, we'll walk through each section of the AI summary, along with how that content is generated — so you can interpret the feedback with confidence.

Tip: Not sure how to access the AI summary in Hire? Learn how to request and summarize interviewer feedback here.

How does the AI summary work in PI Hire?

The AI summary in Hire incorporates the following data:

Once a user clicks "Generate AI summary," all of the above will be taken into account when summarizing the findings.

How is the AI summary presented?

The AI summary consists of 3 key sections:

  • Interview Highlights

  • The Interview Summary

  • An Assessment vs. Interview Comparison

The first 2 sections are fairly straightforward. Interview Highlights shows high-level feedback from each interviewer, so you can quickly see each person's overall thoughts on a candidate.

The Interview Summary offers a more detailed recap (and synthesis) of how each interviewer felt about the candidate. This section also includes a visual of summarized strengths and potential cautions, meant as a quick visual reference to help gauge candidate fit.

The Assessment vs. Interview Comparison, meanwhile, is a more complex summary that incorporates not only interviewer feedback, but also the job's Behavioral Target and candidate's Behavioral Assessment results.

This section is designed to contextualize your interview team's feedback in relationship to the candidate's PI behavioral data (and the Job Target at large).

Example: If the interviewer feedback for a candidate aligns with the candidate's Behavioral Assessment data, it indicates that the candidate's in-person behaviors manifested similar to how they described themself on paper.

If there's misalignment between the assessment and the interview, it may signal a need to further confirm a candidate's strengths (or probe for potential cautions).

FAQs

Is it possible to regenerate the AI summary?

Yes. You can regenerate the summary, provided you receive additional interviewer feedback for the AI to incorporate.

The "Generate AI summary" button is greyed out. Why?

There are a few reasons why the "Generate AI summary" button may be greyed out.

  • Not enough interviewers have provided feedback. PI Hire requires at least 2 sets of interviewer notes to generate an AI summary.

  • The content was too brief to summarize. If you have 2 sets of notes, but the button is still greyed out, it likely means the content was too brief to summarize. We recommend gathering additional feedback from other interviewers before trying to summarize.

  • You've already summarized the data. In this case, you won't be able to generate a new summary unless a new interviewer submits feedback.

Does the AI summary support non-English languages?

Currently, our AI model is English-only. If your organization uses PI in a language other than English, you will not have access to the AI summary feature.

Can the AI summary in Hire be disabled?

The use of the interviewer feedback AI summary can be disabled at any time by an organization admin.

To disable the summary feature (org admins only):

  1. Log in to the PI software.

  2. Click the gear icon in the top-right corner.

  3. You will land on the Administration page.

  4. Click the "Organization Settings" tab.

  5. Select "AI Settings" on the left-hand side.

  6. Deselect the checkbox titled, "Enable use of AI to summarize interview feedback for this organization."

  7. Click "Save Changes" to confirm.

Note: Disabling the AI summary does not retroactively disable past summaries. Hire users with access to those summaries will still be able to view them.

Additional support

Applied learning

Did this answer your question?