AI Interview Chatbot

Table of Contents

• Overview
• Technical Aspect
• Installation
• Usage
• Configuration
• Examples
• Directory Tree
• Troubleshooting
• Bug / Feature Request
• Technologies Used

Overview

This project is an AI Interview Chatbot that uses OpenAI's GPT models (GPT-4o-mini by default) and the Streamlit framework. The chatbot generates job-specific interview questions and evaluates candidate responses using advanced language model capabilities.

Interview-bot-Video.mp4

Technical Aspect

The Interview Chatbot project consists of three main components:

1. Question Generation: Creates job-specific interview questions based on a configurable job description
2. Response Collection: Interactive chat interface for collecting candidate answers
3. Evaluation: AI-powered assessment of candidate responses against job requirements

The project uses OpenAI's API with support for custom base URLs, making it compatible with OpenAI-compatible endpoints.

Installation

Prerequisites

• Python 3.11 or higher
• uv package manager (recommended) or pip

Setup Environment Variables

1. Copy the .env.example file to create your own .env file:

   cp .env.example .env

2. Choose your LLM provider - Edit the .env file:

   Option A: Using OpenAI

   LLM_PROVIDER=openai
   OPENAI_API_KEY=your_actual_api_key_here
   OPENAI_BASE_URL=https://api.openai.com/v1
   OPENAI_MODEL=gpt-4o-mini

   • Replace your_actual_api_key_here with your actual OpenAI API key
   • Update OPENAI_BASE_URL if using a custom OpenAI-compatible endpoint
   • Change OPENAI_MODEL to use a different model (e.g., gpt-4, gpt-4-turbo)

   Option B: Using Ollama (Local/Free)

   LLM_PROVIDER=ollama
   OLLAMA_BASE_URL=http://localhost:11434/v1
   OLLAMA_MODEL=llama3.2
   OLLAMA_API_KEY=ollama

   • Install Ollama from ollama.ai
   • Pull a model: ollama pull llama3.2 (or llama3.1, mistral, qwen2.5, etc.)
   • Start Ollama: ollama serve (usually runs automatically)
   • Change OLLAMA_MODEL to any model you have pulled

Installation Methods

Method 1: Using uv (Recommended)

# Install uv if you haven't already
curl -LsSf https://astral.sh/uv/install.sh | sh

# Clone the repository
git clone https://github.com/nehalvaghasiya/interview-bot.git
cd interview-bot

# Install dependencies using uv
uv sync

# Activate the virtual environment
source .venv/bin/activate  # Linux/Mac
# or
.venv\Scripts\activate  # Windows

# Run the application
streamlit run chatbot.py

Method 2: Using uv with requirements.txt

# Create a virtual environment and install dependencies
uv venv
source .venv/bin/activate  # Linux/Mac
uv pip install -r requirements.txt

# Run the application
streamlit run chatbot.py

Method 3: Using pip

# Create a virtual environment
python3 -m venv .venv
source .venv/bin/activate  # Linux/Mac
# or
.venv\Scripts\activate  # Windows

# Install dependencies
pip install -r requirements.txt

# Run the application
streamlit run chatbot.py

Usage

1. Start the application:

   streamlit run chatbot.py

2. Access the chatbot: Open your browser at http://localhost:8501

3. Interact with the bot:

   • The chatbot will greet you and begin asking interview questions
   • Type your answers in the text input field
   • Press Enter to submit each answer
   • After all questions are answered, you'll receive an AI-generated evaluation

Configuration

Customizing the Job Description

Edit config.py to customize the job description and prompts:

class Parameters:
    MODEL = os.environ.get("OPENAI_MODEL", "gpt-4o-mini")
    
    JOB_DESCRIPTION = """
    Your custom job description here...
    """
    
    # Customize prompts as needed
    QUESTIONS_PROMPT = "..."
    EVALUATION_PROMPT = "..."

Environment Variables

Variable	Description	Default	Required
LLM_PROVIDER	LLM provider to use (openai or ollama)	openai	Yes
OpenAI Settings
OPENAI_API_KEY	Your OpenAI API key	-	If using OpenAI
OPENAI_BASE_URL	OpenAI API base URL	https://api.openai.com/v1	No
OPENAI_MODEL	OpenAI model name	gpt-4o-mini	No
Ollama Settings
OLLAMA_BASE_URL	Ollama API base URL	http://localhost:11434/v1	No
OLLAMA_MODEL	Ollama model name	llama3.2	No
OLLAMA_API_KEY	Placeholder (Ollama doesn't need real key)	ollama	No

Examples

Example Interview Flow

1. Bot: "Hello! I'm your interviewer bot powered by OpenAI. I will ask you a few questions, and your responses will be evaluated. Let's get started."

2. Bot: "What is your experience with Machine Learning and AI implementation?"

3. User: "I have 5 years of experience implementing ML models in production..."

4. Bot: "Can you describe a challenging NLP project you've worked on?"

5. User: "I developed a sentiment analysis system for customer feedback..."

6. (After all questions)

7. Bot: "Thank you for your thoughtful responses. Based on your answers, it appears that your skills, experience, and understanding align well with the requirements of the role..."

Directory Tree

interview-bot/
├── .env.example              # Environment variables template
├── .github/                  # GitHub configuration
├── .gitignore               # Git ignore rules
├── README.md                # This file
├── chatbot.py               # Main Streamlit application
├── config.py                # Configuration and prompts
├── utils.py                 # Utility functions (API calls, text processing)
├── pyproject.toml           # Project metadata and dependencies (uv)
├── requirements.txt         # Python dependencies (pip)
├── uv.lock                  # Locked dependencies (uv)
├── devtools/                # Development tools
│   └── lint.py
├── images/                  # Project images
│   ├── openai.png
│   └── streamlit.jpg
├── src/                     # Source package (if using as package)
│   └── interview_bot.egg-info/
└── tests/                   # Test files
    └── test_placeholder.py

Troubleshooting

Common Issues

Module Not Found Errors

If you encounter import errors:

# Ensure you're in the virtual environment
source .venv/bin/activate  # Linux/Mac

# Reinstall dependencies
uv pip install -r requirements.txt
# or
pip install -r requirements.txt

API Connection Errors

For OpenAI:

• Verify your .env file exists and contains valid credentials
• Check that OPENAI_API_KEY is set correctly
• Ensure OPENAI_BASE_URL is accessible from your network

For Ollama:

• Ensure Ollama is running: ollama serve (or check if it's running as a service)
• Verify the model is installed: ollama list
• Pull the model if missing: ollama pull llama3.2
• Check Ollama is accessible: curl http://localhost:11434/api/tags
• Verify OLLAMA_BASE_URL points to the correct endpoint

Model Not Found Error

For OpenAI:

• Update the OPENAI_MODEL in your .env file to a model available in your account
• Common models: gpt-4o-mini, gpt-4o, gpt-4-turbo, gpt-4

For Ollama:

• List available models: ollama list
• Pull the desired model: ollama pull <model-name>
• Popular models: llama3.2, llama3.1, mistral, qwen2.5, phi3
• Update OLLAMA_MODEL in .env to match an installed model

Ollama-Specific Issues

Ollama not responding:

# Check if Ollama is running
ps aux | grep ollama

# Start Ollama
ollama serve

Slow responses with Ollama:

• Local models require sufficient RAM and compute
• Consider using smaller models: phi3, llama3.2:1b
• Or faster models: qwen2.5:3b, mistral

Streamlit Errors

If Streamlit throws errors:

# Clear Streamlit cache
streamlit cache clear

# Restart the application
streamlit run chatbot.py

Getting Help

If issues persist:

1. Check that all dependencies are correctly installed
2. Verify Python version is 3.11 or higher: python --version
3. Review error logs in the terminal
4. Open an issue on GitHub with error details

Bug / Feature Request

If you find a bug or the chatbot doesn't work as expected, please open an issue here with:

• Description of the issue
• Steps to reproduce
• Error messages or screenshots
• Your environment details (OS, Python version)

If you'd like to request a new feature, open an issue here with:

• Feature description
• Use case and benefits
• Example scenarios

Technologies Used

• OpenAI API - GPT-4o-mini and other models for question generation and evaluation
• Ollama - Local LLM support (Llama 3.2, Mistral, Qwen, and more)
• Streamlit - Web framework for the interactive chat interface
• Python-dotenv - Environment variable management
• uv - Fast Python package manager

Supported LLM Providers

OpenAI

• Cloud-based API
• High-quality responses
• Requires API key and costs per token
• Models: GPT-4o, GPT-4o-mini, GPT-4-turbo, GPT-3.5-turbo

Ollama

• Run models locally on your machine
• Free and private
• No internet required after model download
• Popular models: Llama 3.2, Llama 3.1, Mistral, Qwen 2.5, Phi-3
• Learn more: ollama.ai