MillerKnoll AI
Executive Summary

MK Comment Signal

Open-text survey comments are the richest — and most unread — part of an engagement survey, because reading and coding thousands of them by hand isn’t realistic. MK Comment Signal turns them into ranked themes with real verbatim evidence, PII removed, ready to inform an organizational decision.

By the numbers

What it delivers

4–5 themes
Per survey question, frequency-ranked
≥ 12 quotes
Verbatim evidence per theme
PII-redacted
Names, emails, IDs removed automatically
Jan 20, 2026
In production for HR / People
Why it works

The principles

01

Evidence, not just labels

Every theme carries at least a dozen verbatim quotes ranked by frequency — so a leader sees the actual words behind the finding, not a black-box summary.

02

Privacy built in

Names, emails, and employee IDs are redacted automatically, so sensitive open-text feedback can be analyzed and shared without exposing individuals.

03

Structured for a decision

Themes are ordered by support count and exported to a formatted Word document — the analysis lands as something a People team can act on, not raw output.

Where it stands

MK Comment Signal has been in Production since January 20, 2026, on GPT-5 Chat, used by the HR / People Analytics team. It extracts themes from Perceptyx survey comments across four questions, ranks them by frequency with a dozen-plus verbatim quotes each, redacts PII automatically, and can export a formatted Word document to OneDrive.