Manual Evaluation allows teams to perform manual quality assessments of interactions directly inside Birdie for criteria that cannot be reliably monitored by AI.
The feature is designed to centralize all manual evaluation workflows, ensuring governance, traceability, and consistency, while integrating manual results with AI-driven metrics. All manual evaluations start at the Area level, providing a scalable and standardized entry point that represents the full operational scope.
Manual Evaluation helps Birdie remain the single source of truth for quality monitoring by unifying configuration, execution, and analytics in one place.
Viewing & Searching
Manual evaluation results are available throughout Birdie analytics and dashboards, alongside AI-driven evaluations.
Filtering & Exploration
Users can filter and explore evaluation data by:
Evaluation source (Manual or AI)
All views support drill-down from high-level metrics to individual evaluations, ensuring complete transparency.
Creating New Data
Manual Evaluation involves two different actions: configuring manual criteria and performing a manual evaluation.
This article focuses on how to use Manual Evaluation. Detailed configuration of criteria is covered in a separate article.
1. Creating Manual Criteria
Before performing manual evaluations, manual criteria must be configured by an Admin.
If you need help about how to create a criterion please check Criteria
Once criteria are configured, you can perform manual evaluations directly from an Area.
Go to the desired Area in Birdie
Click Start Manual Evaluation
Select one or more Reasons to include in the evaluation
Birdie automatically opens an interaction for evaluation
Complete the evaluation form
Add optional comments or reasoning (if applicable)
After submission, the evaluation is automatically linked to the Area, Reason, agent, and interaction, and becomes available in dashboards and reports.
Modifying Existing Data
What Can Be Modified
Associations between criteria and Reasons
Reason participation within Areas
Statistical proportions used for sampling
Only authorized Admins / Innerloop Specialists can create or modify criteria and configuration
All changes apply only to future evaluations and do not affect historical data.
Manual evaluations cannot be deleted once submitted, ensuring data integrity and compliance.
Supported Actions
Disable criteria so they are no longer used in future evaluations
Remove Reasons from an Area
Deactivate Areas that are no longer active
Historical evaluation data remains available for analytics and auditing purposes.
Troubleshooting & FAQs
No. Forms are dynamically generated based on the criteria associated with the selected Reason to ensure governance and consistency.
Why was a different Reason selected than expected?
Birdie applies statistical sampling rules to ensure representative distribution of evaluations across Reasons.
Viewers can perform manual evaluations
Supervisors have view-only access to results
Admins manage criteria and configuration
