Understanding the answer quality score - Realm Docs

The Answer quality score helps you quickly assess how reliable an AI-generated answer is. The score appears next to each answer in the RFP grid. A detailed breakdown is available in the Answer details panel. The score is based on three factors.

Trustworthiness

Evaluates the quality of the sources used to generate the answer. Sources are ranked in tiers. For example, answers grounded in verified past RFPs or your organization context are considered more reliable than informal sources like Slack conversations.

Completeness

Checks whether the response fully answers the question. Answers that only partially address the question will be flagged.

Sensitivity

Highlights questions that may require human judgment. Topics like pricing, security details, or company-specific information are often marked as sensitive.

Score levels

Scores are shown on a four-level scale: Very High, High, Medium, and Low. Lower scores indicate answers that should be reviewed more carefully.