MillerKnoll AI
Executive Summary

Survey Insights Analyzer

A West Michigan client visit is a high-stakes moment in the sales process, and every visit is surveyed — but the feedback sat in spreadsheets no one had time to fully read. Survey Insights Analyzer turns that WMCE data into an executive-quality read: named themes, dimension-level trends, NPS, and a cited summary, grounded only in the survey record.

By the numbers

What it delivers

Sonnet 4.6
The model behind the analysis
3 sources
SharePoint + 2 Excel files — web search off
~13 dimensions
From SME sessions to transportation & NPS
POC in TEST
For Erica Schenck & Client Experience review
Why it works

The principles

01

Grounded, and only grounded

The agent reads three vetted sources and nothing else — web search is disabled — so every theme and number traces back to the actual WMCE survey record, not the model’s general knowledge.

02

Quantify and cite, never fabricate

Findings are backed by specific scores, response counts, and percentages with citations; where the data can’t support a claim, the agent says so rather than inventing one.

03

Recency and relevance discipline

COVID-era virtual visits are excluded entirely, and any insight drawn from data older than 24 months is flagged — so leadership acts on a current, representative picture.

Where it stands

Survey Insights Analyzer is a POC deployed to the Copilot Agents TEST environment on Claude Sonnet 4.6, provisioned for Erica Schenck and the Client Experience team to review. It’s grounded exclusively in the WMCESurveys SharePoint library and two Excel files, extracts themes and dimension-level trends across the WMCE survey, and drafts follow-up emails via the built-in action. UAT with the Client Experience team and security-group definition are the next steps toward promotion.