Rafiki AI provides a comprehensive set of coaching tools that help you identify what top performers do differently, share those practices with the team, and track coaching engagement. Here is how to use each tool.

Scorecards give managers a consistent framework for evaluating calls. Create templates with weighted questions in Settings > Scorecards, then score individual calls on a 0-5 scale per question. Rafiki AI calculates weighted averages automatically. Use Scorecard Metrics in Analytics to view heatmaps showing how each rep performs across every scoring dimension, making it easy to spot patterns and target coaching.

Build playlists of meeting recordings and snippets that demonstrate best practices -- or common mistakes to avoid. Share playlists with your team so reps can learn from real conversations. Create separate playlists for different skills: discovery calls, objection handling, closing techniques, product demos, and more.

The Coaching dashboard tracks how actively managers are coaching their teams. Monitor metrics including calls listened to, meetings attended, comments left, and scorecards completed per team member. Filter by date range and drill down to specific meetings. This ensures coaching is happening consistently across the organization.

Role Play lets reps practice conversations with AI-powered buyer personas before getting on real calls. Configure personas with different personalities, objection styles, and difficulty levels (easy, medium, hard). After each session, the AI scores performance across five dimensions: discovery questions, objection handling, value proposition, closing technique, and active listening.

Use Analytics > Interaction Metrics to compare each rep's conversation patterns against recommended ranges:

Scatter plots show where each rep falls relative to these benchmarks. Topic Stats provide three-level drill-down: organization average, per-rep breakdown, and rep vs. team comparison.

Leave comments on specific portions of meeting transcripts to provide targeted feedback. Use @ mentions to notify team members directly. Comments are linked to the exact moment in the conversation, making feedback specific and actionable rather than general.