- Navigate to
/register - Enter email and password
- Verify email (if required)
- Log in at
/login
- Go to Resumes → Upload Resume
- Upload PDF resume
- Wait for analysis (30-60 seconds)
- Review extracted data
Supported Formats: PDF only
- Go to Interviews → New Interview
- Select analyzed resume
- (Optional) Add job description
- Click Start Interview
graph LR
A[Start] --> B[Greeting]
B --> C[Questions]
C --> D{User Response}
D -->|Answer| C
D -->|Code Request| E[Sandbox]
E --> F[Code Review]
F --> C
D -->|Complete| G[Evaluation]
G --> H[Feedback]
H --> I[End]
The orchestrator manages flow through phases: intro (greeting), exploration (resume-based questions), technical (code exercises), and closing (evaluation). The LLM decides transitions based on conversation context, user responses, and interview progress. Code submissions trigger immediate routing to code_review, bypassing normal question flow. The agent guides you to the sandbox when appropriate, and code execution results inform the next questions.
| Phase | Description | Duration |
|---|---|---|
| Intro | Greeting, initial questions | 2-5 min |
| Exploration | Resume-based questions | 10-20 min |
| Technical | Code exercises, problem-solving | 10-15 min |
| Closing | Final questions, wrap-up | 2-5 min |
- During interview, agent may guide you to sandbox
- Sandbox appears on right side of screen
- Select language (Python/JavaScript)
- Type code in editor
- Click Run to test locally
- Click Submit to submit to interview
- Agent reviews code and provides feedback
| Language | Version | Features |
|---|---|---|
| Python | 3.11 | Full standard library, no external packages |
| JavaScript | Node 18 | Standard Node.js APIs |
What happens:
- Code executed in isolated Docker container
- Execution results analyzed
- Code quality assessed (correctness, efficiency, readability)
- Agent provides feedback
Best Practices:
- Write clear, readable code
- Add comments for complex logic
- Test your code before submitting
- Follow language best practices
- Speak clearly: Enunciate words, avoid mumbling
- Be concise: Answer directly, avoid rambling
- Ask for clarification: If question unclear, ask agent to rephrase
- Think aloud: Explain your thought process
- Start simple: Begin with basic solution, then optimize
- Admit uncertainty: It's okay to say "I'm not sure, but I think..."
- Read carefully: Understand problem before coding
- Test edge cases: Consider boundary conditions
- Refactor if needed: Improve code after initial solution
| Skill | Weight | Description |
|---|---|---|
| Communication | 25% | Clarity, articulation, engagement |
| Technical Knowledge | 30% | Depth, accuracy, expertise |
| Problem-Solving | 25% | Approach, logic, creativity |
| Code Quality | 20% | Correctness, efficiency, readability |
- Overall Score: Weighted average (0.0-1.0)
- Skill Scores: Individual scores per dimension
- Strengths: 2-3 things you did well
- Weaknesses: 2-3 areas for improvement
- Recommendations: 3-5 actionable next steps
| Score Range | Interpretation |
|---|---|
| 0.8-1.0 | Excellent - Strong performance |
| 0.6-0.79 | Good - Solid with minor gaps |
| 0.4-0.59 | Fair - Needs improvement |
| 0.0-0.39 | Poor - Significant gaps |
No sound:
- Check browser permissions (microphone/speaker)
- Verify internet connection
- Refresh page and reconnect
Echo/feedback:
- Use headphones
- Reduce speaker volume
- Check microphone sensitivity
Agent disconnected:
- Refresh page
- Check internet connection
- Restart interview if needed
Slow responses:
- Check network speed
- Close other browser tabs
- Wait for agent to respond (may take 5-10 seconds)
Code won't run:
- Check syntax errors
- Verify language selected correctly
- Review error messages
Timeout:
- Code execution limited to 30 seconds
- Optimize slow algorithms
- Break into smaller functions
- Prepare: Review your resume before interview
- Practice: Do mock interviews to get comfortable
- Focus: Minimize distractions during interview
- Take notes: Jot down key points during conversation
- Review feedback: Learn from each interview
Need help?
- Check FAQ (coming soon)
- Email: support@interviewlab.com
- Discord: Join community