How to Auto-Score Candidates with AI

Automatically evaluate candidates with AI when they reach specific pipeline stages.

How It Works

When a candidate enters a stage with AI scoring automation:

1. System triggers AI evaluation
2. AI analyzes candidate against your criteria
3. Score appears in candidate profile
4. Optional: Move candidate based on score

Setting Up Auto AI Scoring

Prerequisites

1. Set up scoring criteria on job's AI Scoring tab
2. Have sufficient AI credits in your account

Create the Automation

1. Open a job
2. Go to the Workflow tab
3. Click the bolt icon on the target stage
4. Click Add Automation
5. Select AI Score
6. Configure options

Configuration Options

Move If Score Above

Set a threshold score. Candidates scoring above this are automatically moved to a specified stage.

Example: Score above 70 → Move to "Phone Screen"

Disqualify If Score Below

Set a minimum threshold. Candidates below this are automatically disqualified.

Example: Score below 40 → Disqualify

Both Options Together

Create automatic routing:

• Score 70+ → Move to "Interview"
• Score 40-69 → Stay for manual review
• Score below 40 → Disqualify

Example Automation Flows

High-Volume Screening

Stage: Applied

• Auto AI Score
• Score 60+ → Move to "Qualified"
• Score below 40 → Disqualify
• Purpose: Filter large applicant pools quickly

Post-Questionnaire Evaluation

Stage: Questionnaire Complete

• Auto AI Score (uses questionnaire data)
• Score 75+ → Move to "Phone Screen"
• Purpose: Evaluate detailed responses

Two-Phase Scoring

Stage 1: Applied

• Standard model AI Score
• Score 50+ → Move to "Under Review"

Stage 2: Under Review

• Premium model AI Score (deeper analysis)
• Score 70+ → Move to "Interview"

Best Practices

Pre-Screen First

Save AI credits by pre-screening:

1. Use knockout questions to eliminate unqualified
2. Only AI score candidates who pass basics

Set Appropriate Thresholds

• Too high: Missing good candidates
• Too low: Not filtering effectively
• Start conservative, adjust based on results

Combine with Manual Review

AI scoring assists, not replaces, human judgment:

• Review borderline scores manually
• Calibrate AI criteria based on outcomes
• Use for prioritization, not final decisions

Credit Considerations

• Each AI score uses credits from your plan
• Premium model uses ~10x more credits
• Monitor usage in Settings / Billing
• Pre-screen to reduce unnecessary scoring

Updated on: 22/12/2025