Skip to content

Latest commit

 

History

History
200 lines (139 loc) · 5.42 KB

File metadata and controls

200 lines (139 loc) · 5.42 KB

User Guide

Getting Started

1. Create Account

  1. Navigate to /register
  2. Enter email and password
  3. Verify email (if required)
  4. Log in at /login

2. Upload Resume

  1. Go to ResumesUpload Resume
  2. Upload PDF resume
  3. Wait for analysis (30-60 seconds)
  4. Review extracted data

Supported Formats: PDF only

3. Start Interview

  1. Go to InterviewsNew Interview
  2. Select analyzed resume
  3. (Optional) Add job description
  4. Click Start Interview

Interview Flow

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]
Loading

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.

Phases

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

Using the Code Sandbox

Accessing Sandbox

  1. During interview, agent may guide you to sandbox
  2. Sandbox appears on right side of screen
  3. Select language (Python/JavaScript)

Writing Code

  1. Type code in editor
  2. Click Run to test locally
  3. Click Submit to submit to interview
  4. Agent reviews code and provides feedback

Supported Languages

Language Version Features
Python 3.11 Full standard library, no external packages
JavaScript Node 18 Standard Node.js APIs

Code Submission

What happens:

  1. Code executed in isolated Docker container
  2. Execution results analyzed
  3. Code quality assessed (correctness, efficiency, readability)
  4. Agent provides feedback

Best Practices:

  • Write clear, readable code
  • Add comments for complex logic
  • Test your code before submitting
  • Follow language best practices

Interview Tips

Communication

  • Speak clearly: Enunciate words, avoid mumbling
  • Be concise: Answer directly, avoid rambling
  • Ask for clarification: If question unclear, ask agent to rephrase

Technical Questions

  • 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..."

Code Exercises

  • Read carefully: Understand problem before coding
  • Test edge cases: Consider boundary conditions
  • Refactor if needed: Improve code after initial solution

Understanding Feedback

Skill Breakdown

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

Feedback Components

  1. Overall Score: Weighted average (0.0-1.0)
  2. Skill Scores: Individual scores per dimension
  3. Strengths: 2-3 things you did well
  4. Weaknesses: 2-3 areas for improvement
  5. Recommendations: 3-5 actionable next steps

Interpreting Scores

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

Common Issues

Audio Problems

No sound:

  • Check browser permissions (microphone/speaker)
  • Verify internet connection
  • Refresh page and reconnect

Echo/feedback:

  • Use headphones
  • Reduce speaker volume
  • Check microphone sensitivity

Connection Issues

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 Execution Errors

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

Best Practices

  1. Prepare: Review your resume before interview
  2. Practice: Do mock interviews to get comfortable
  3. Focus: Minimize distractions during interview
  4. Take notes: Jot down key points during conversation
  5. Review feedback: Learn from each interview

Support

Need help?