AuthentiVoice’s review system enables teams to collaborate efficiently on call analysis, with role-based workflows, progress tracking, and consensus building.

Overview

The multi-reviewer workflow system is designed for organizations that need multiple perspectives on call analysis:

Role-Based Access

Admin, Supervisor, and Reviewer roles

Smart Assignment

Intelligent workload distribution

Progress Tracking

Real-time visibility into review status

Quality Scoring

Comprehensive scoring system

User Roles & Permissions

Role Hierarchy

  • Admin
  • Supervisor
  • Reviewer
Full system access including:
  • User management
  • Organization settings
  • All supervisor permissions
  • System configuration
  • Integration management
  • Audit logs access

Review Assignment

Manual Assignment

Supervisors can manually assign reviews based on:
1

Individual Assignment

Assign specific calls to individual reviewers
await assignReview({
  analysisId: "123",
  reviewerId: "reviewer-456",
  priority: "high",
  dueDate: "2024-01-15"
});
2

Bulk Assignment

Distribute multiple calls across team members
await bulkAssignReviews({
  analysisIds: ["123", "124", "125"],
  reviewerIds: ["reviewer-456", "reviewer-789"],
  distributionMethod: "round-robin"
});
3

Smart Assignment

AI-powered assignment based on expertise and workload
await smartAssignReviews({
  analysisIds: ["123", "124", "125"],
  criteria: {
    balanceWorkload: true,
    matchExpertise: true,
    respectAvailability: true
  }
});

Assignment Strategies

Evenly distributes calls among available reviewersBest for: Teams with similar skill levelsConfiguration:
{
  "method": "round-robin",
  "skipUnavailable": true,
  "maxPerReviewer": 10
}
Assigns based on current reviewer workloadBest for: Maintaining balanced productivityConfiguration:
{
  "method": "workload",
  "targetCallsPerDay": 20,
  "considerComplexity": true
}
Matches calls to reviewer expertiseBest for: Specialized review requirementsConfiguration:
{
  "method": "expertise",
  "categories": ["sales", "support", "compliance"],
  "fallbackMethod": "round-robin"
}

Review Process

Review Workflow

Review Interface

The review interface is designed for efficiency, allowing reviewers to quickly assess calls while maintaining thoroughness.
Key features of the review interface:
  1. Synchronized Playback: Audio plays alongside transcript highlighting
  2. AI Insights Panel: View fraud indicators and red flags
  3. Scoring Widgets: Quick scoring for different criteria
  4. Comment System: Add timestamped comments
  5. Keyboard Shortcuts: Speed up review process

Scoring System

  • Default Scorecard
  • Custom Scorecards
interface ReviewScorecard {
  overallScore: number;        // 1-10
  categories: {
    compliance: number;        // 1-10
    quality: number;          // 1-10
    accuracy: number;         // 1-10
    professionalism: number;  // 1-10
  };
  fraudAssessment: {
    agreesWithAI: boolean;
    confidenceLevel: 'low' | 'medium' | 'high';
    additionalFlags: string[];
  };
}

Progress Tracking

Team Dashboard

Monitor team performance in real-time:

Reviews Pending

Track unassigned and in-progress reviews

Completion Rate

Monitor daily/weekly completion metrics

Quality Scores

Track average quality scores by reviewer

Individual Metrics

Track reviewer performance:
MetricDescriptionTarget
Reviews/DayAverage daily review completion20-30
Accuracy RateAgreement with consensus/supervisor>90%
Avg Review TimeTime spent per review3-5 min
Quality ScoreAverage score givenBaseline

Multi-Reviewer Consensus

For high-stakes calls, multiple reviewers can assess the same recording:
1

Configure Consensus Rules

const consensusConfig = {
  minReviewers: 3,
  requiredAgreement: 0.8, // 80% agreement
  tieBreaker: 'supervisor'
};
2

Assign Multiple Reviewers

await assignConsensusReview({
  analysisId: "high-risk-123",
  reviewerIds: ["r1", "r2", "r3"],
  blindReview: true // Reviewers can't see others' scores
});
3

Calculate Consensus

System automatically calculates consensus based on:
  • Majority agreement on key indicators
  • Average scores within threshold
  • Supervisor override if needed

Notifications & Alerts

Review Notifications

  • Email Notifications
  • In-App Alerts
  • Webhook Integration
  • New review assignments
  • Due date reminders
  • Escalation alerts
  • Daily summary reports

Best Practices

Following these best practices ensures consistent, high-quality reviews across your team.

For Supervisors

  1. Balance Workloads: Monitor reviewer capacity and adjust assignments
  2. Set Clear Expectations: Define quality standards and timelines
  3. Regular Calibration: Conduct team calibration sessions
  4. Provide Feedback: Regular performance reviews and coaching

For Reviewers

  1. Maintain Consistency: Apply scoring criteria uniformly
  2. Document Thoroughly: Add clear comments for decisions
  3. Flag Uncertainties: Escalate unclear cases to supervisors
  4. Meet Deadlines: Complete reviews within SLA timeframes

API Integration

Integrate the review system with your existing tools:
# Create a review assignment
def assign_review(analysis_id, reviewer_id):
    response = requests.post(
        f"{API_BASE_URL}/reviews/assign",
        json={
            "analysisId": analysis_id,
            "reviewerId": reviewer_id,
            "priority": "normal",
            "dueDate": datetime.now() + timedelta(days=1)
        },
        headers={"X-API-Key": API_KEY}
    )
    return response.json()

# Submit a review
def submit_review(review_id, scores):
    response = requests.put(
        f"{API_BASE_URL}/reviews/{review_id}",
        json={
            "scores": scores,
            "status": "completed",
            "comments": "Review completed successfully"
        },
        headers={"X-API-Key": API_KEY}
    )
    return response.json()

Reporting & Analytics

Generate comprehensive reports on review operations:
  • Review completion rates by team member
  • Average review times
  • Quality score distributions
  • Workload analysis
  • Inter-rater reliability scores
  • Consensus agreement rates
  • Escalation patterns
  • Training needs identification
  • Total reviews completed
  • Average turnaround time
  • SLA compliance rate
  • Resource utilization

Next Steps