Expression Measurement for Emotional Intelligence | Hume Blog | Hume AI

About IRIS Insights

IRIS Insights (formerly EarningsEdge AI) is an AI-powered conversational intelligence platform helping organizations, from contact centers to correctional departments, turn human interactions into actionable emotional insight.

By exclusively integrating Hume’s Expression Measurement API, IRIS adds a precise emotional layer to its language analysis stack. This allows teams to detect subtle changes in tone like frustration, confusion, or disengagement across every customer or inmate interaction.

In QA workflows, IRIS has seen measurable gains: average CSAT scores now sit at 66/100, and etiquette scores hit a perfect 50/50. Dead air—a marker of low engagement—was tracked at 15%, driving new coaching initiatives to improve flow and emotional continuity.

The Challenge

Traditional sentiment tools lacked the resolution and accuracy to meet IRIS’s needs. They relied on general tonal scoring or binary classifications, which often missed key emotional shifts.

IRIS needed an API that could deliver fine-grained emotional data, capturing dozens of distinct expressions moment by moment.

The Solution: Precise Emotion Detection

Hume’s Expression Measurement API offered the emotional depth IRIS was looking for:

• Word-by-Word Emotion Detection: Unlike tools that deliver vague sentiment, the API outputs 48 distinct emotions per utterance including empathy, anxiety, frustration, and relief.
• Emotion-Based Prioritization: CX and QA teams use emotional spikes to flag calls worth reviewing and intervene proactively.
• Seamless Integration: With a clean, well-documented API and easy-to-use output format, integration into IRIS’s core analytics engine required minimal engineering effort.

Emotional Intelligence as a Differentiator

Since embedding Hume’s Expression Measurement API, IRIS clients have experienced:

• Clearer Visibility into Sentiment: Emotional signals like rising anxiety or disengagement are now measurable and actionable.

“We were blind to what our customers were feeling until we had emotion as a signal. Now we coach smarter and intervene faster,” shared a CX Manager at one partner company.

• Improved Agent Performance: Coaching is more personalized, grounded in emotion-driven feedback instead of static transcripts. Coaching programs now use emotion trends to target specific skills—such as objection handling, recap, and probing questions—which previously scored between 24–64% across QA reviews.
• Market Differentiation: Early customers cite emotion analysis as a unique value-add that legacy QA and sentiment tools fail to deliver.
• A major milestone: IRIS earned regulatory approval to pilot its correctional platform—validating the importance of emotional intelligence in public safety contexts.

For more information on how empathic AI can enhance your digital solutions, please contact Hume AI.