AI Interview

What is AI Interview? #

AI Interview is an automated grading feature that evaluates participant responses in interview modules. When enabled, our AI system automatically processes and scores interview answers, providing a consistent and objective evaluation based on your specified criteria.

How does AI Interview work? #

The AI Interview feature automatically grades interview module responses when participants submit their answers. The system analyzes the correlation between questions and answers, providing scores on a scale of 0-100 based on your chosen scoring method.

How do I create an AI Interview module? #

Follow these step-by-step instructions to set up your AI Interview:

Step 1: Access Modules #

Go to Modules in your Algobash dashboard

Step 2: Create New Interview Module #

Click “+ New Module” > Select “Interview”

Step 3: Configure Basic Settings #

  1. Add your module title – Choose a descriptive name for your interview module
  2. Fill in module duration – Set the total time allowed for the entire interview

Step 4: Choose Expected Language ⚠️ CRITICAL STEP #

Select your expected language carefully. Our AI will specifically expect answers in the chosen language and process responses based on this selection.

Important Warning: If you choose English but participants answer in another language (or vice versa), the AI grading will fail. Ensure language consistency between your selection, questions, and expected participant responses.

Step 5: Add Questions by Language #

Add your questions based on the selected language. If you choose multiple languages, make sure to add questions in each corresponding language properly.

Step 6: Set Answer Duration #

Choose how long participants are allowed to answer each question:

  • Minimum: 1 minute per question
  • Maximum: 4 minutes per question

Step 7: Configure AI Scoring Method #

Choose from four AI scoring options:

1. None #

  • Use case: Manual grading or no AI involvement
  • Function: Disables AI grading entirely

2. Default #

  • Use case: Standard AI evaluation
  • Function: Gives AI full control of scoring
  • Output: Correlation score between answers and questions on a scale of 0-100

3. Hints #

  • Use case: Specific answer validation
  • Function: If the answer is similar to one of the hints, then return 100, otherwise return 0
  • Output: Binary scoring (0 or 100) based on hint matching

4. Custom Prompt #

  • Use case: Specialized evaluation criteria
  • Function: Score the correlation between the answer and the question on a scale of 0-100 based on the suggested prompt
  • Tip: To create effective AI prompts, reference our article on AI prompt best practices
  • Output: Custom scoring based on your specific prompt instructions

What results does AI Interview provide? #

The AI Interview system returns a numerical score between 0-100 for each question. This scoring provides:

  • Consistent evaluation across all participants
  • Objective assessment based on your chosen criteria
  • Immediate results upon participant submission

What should I do if I get unexpected results? #

If the AI grading produces unexpected or inaccurate results, follow this troubleshooting process:

Step 1: Verify Your Setup #

  1. Check the language settings – Ensure the selected language matches your questions and expected participant responses
  2. Review custom prompts – Verify that custom prompts are clear and specific
  3. Confirm question clarity – Make sure your questions are well-written and unambiguous

Step 2: Use the Recalculation Option #

After reviewing and correcting any issues, you have one opportunity to recalculate the AI grading for that assessment.

Step 3: Contact Support #

If issues persist after recalculation, please contact our support team for assistance.

Best Practices for AI Interview Success #

Language Consistency #

  • Always match your language selection with your questions and expected answers
  • If conducting multilingual interviews, clearly separate language-specific sections
  • Inform participants about the expected response language

Question Design #

  • Write clear, specific questions that have definitive answers
  • Avoid ambiguous or overly complex questions for AI grading
  • Test your questions with the AI system before launching to participants

Scoring Method Selection #

  • Use the Default for general correlation-based scoring
  • Use Hints when you have specific correct answers or key points
  • Use Custom Prompt for specialized evaluation criteria
  • Choose None if you prefer manual review

Duration Settings #

  • Allow sufficient time for thoughtful responses (minimum 1 minute)
  • Consider question complexity when setting maximum duration (up to 4 minutes)
  • Balance thorough answers with practical time constraints

Technical Requirements #

  • Stable internet connection for real-time AI processing
  • Sufficient AI credits for the number of questions and participants
  • Compatible browser for optimal interview module functionality

Troubleshooting Common Issues #

AI Grading Fails #

  • Most common cause: Language mismatch between settings and responses
  • Solution: Verify language consistency and recalculate if needed

Low or Unexpected Scores #

  • Possible causes: Unclear questions, inappropriate scoring method, or participant answers don’t match the expected format
  • Solution: Review question clarity and scoring method selection

Technical Errors #

  • If persistent, contact the support team with specific error details and assessment information

For additional technical support or questions about AI Interview functionality, please contact our support team through your Algobash dashboard.