This week’s Tips & Tricks session was led by Reece Lincoln (Chief AI Officer) and Margo Kim (AI Engineer). They introduced the updated Sym Accuracy Score and walked through how it improves fairness, precision, and coaching insights. Additionally, they compared the old system with the new one, explained major enhancements, and provided practical examples. This new Sym Accuracy Score update will benefit both agents and trainers.
Key Takeaways
- Conversation Context Awareness: The new scoring looks at the entire conversation, ensuring agents aren’t penalized for contextually correct answers.
- Multi-Dimensional Scoring: Performance is now broken down into context, intent, and similarity, providing deeper insights and more targeted feedback.
- Evaluation Fairness: The system recognizes effective responses that achieve the right outcome even if worded differently, avoiding rigid script memorization.
- Improved Engagement: Agents are evaluated on their ability to communicate effectively, not just repeat a script, boosting confidence and reducing frustration.
- Better Coaching Insights: The dimensional breakdown allows trainers to pinpoint improvement areas, understand intent, maintain context, or align phrasing.
- Actionable Scoring Data: The new Sym Accuracy Score gives trainers clearer, more reliable performance indicators to design targeted Sym exercises for specific weaknesses.
- Consistency Across Scenarios: Agents performing well with alternative but valid responses are recognized, leading to more accurate performance measurement.
