More Stable Live Audio Sessions & Better Scoring with Context
With version 2.12.0, we’re improving the quality of your live voice training: more stable conversation flow, more reliable scoring, and consistent content in the right languages. You’ll also get an improved help experience and clearer scenario structure.
Highlights
- More stable live voice sessions
- Context for more realistic situations
- Improved scoring logic
- Consistent multilingual experience
- Better help inside the platform
What's inside
More confirmed context for more realistic training
When an organizational context is available, it’s validated and used as a reliable basis in relevant trainings. This makes scenarios feel more consistent and helps feedback match your real-world conditions.
- Validated company context
- Better training embedding
- More coherent feedback
Better conversation observation over time
You get more detailed conversation observations with a clear view of what mattered most and what to focus on next. This makes progress easier to track and helps you sharpen the right skills.
- Clear focus per session
- More insight into your learning path
- Actionable next steps
Consistent language in training and evaluation
The platform ensures that language settings and translations work together properly, with confirmation where needed for your organization. This reduces mismatches between what you say and how you’re evaluated—especially in multilingual teams.
- Fewer language ambiguities
- Confirmation on locale mismatch
- Stable, consistent translations
More reliable scoring for short or complex conversations
Scoring is more robust when a conversation is very short or points can’t be fully covered. You’ll still get a transparent result with references to your transcript and improved coverage of anti-patterns.
- Robust scoring when needed
- Clear reasoning in results
- Consistent anti-pattern coverage
More stable live audio interaction
Live voice interaction is further optimized: responses are more predictable, delays and interruptions are reduced, and the dialog mapping stays stable. The practice feels closer to real phone calls.
- More stable turn-taking
- Less friction while practicing
- Better dialog attribution
Scenario difficulty that’s easier to understand
Each training can now be categorized as EASY, MEDIUM, or HARD. This helps you and your team choose the right training order and build skills with a better progression.
- Difficulty level per scenario
- Better training sequence
- Targeted progression
Customer Avatar Studio for consistent, realistic profiles
When you create or use customer avatars, profiles are derived more consistently from psychological traits. That leads to more realistic counterpart behavior across multiple trainings.
- More consistent avatar profiles
- More realistic counterpart behavior
- Better continuity across sessions
Expanded Help Center for faster support
The Help Center has been expanded with new articles in the user area. You can find answers more quickly and move forward with confidence.
- New help articles
- Easier to find
- Quickly actionable