meeting-insights-analyzer

Meeting transcript analysis identifies speaking ratios, interruptions, filler words, and active listening signals. The Skill processes timestamped, speaker-labeled files to produce summarized patterns and actionable improvement plans.

Timestamped feedback requires timestamped transcripts or speaker-labeled caption files. Processing raw audio or unlabeled text is not supported; speaker identification is mandatory for detecting facilitation behaviors and speaking ratios.

Yes, multiple meeting transcripts can be analyzed together. The Skill compares trends across meetings, such as reductions in interruptions or increases in clarifying questions, to track communication growth over time.

Active listening detection scans speaker-labeled transcripts for verbal signals like clarifying questions and acknowledgments. It outputs timestamped quotes demonstrating these behaviors to support coaching and self-reflection.

Speaking ratio is measured by processing speaker-labeled transcripts to calculate each participant's talk time. The analysis outputs summarized patterns and concrete timestamped examples for professional contexts.

Conflict avoidance detection relies on textual cues like hedging within speaker-labeled transcripts. It cannot interpret non-verbal cues or tone of voice, limiting analysis to explicit verbal patterns captured in the transcript.