🎯 Scout Sense Recruit - Resume Rater

An AI-powered talent matching and interview preparation platform that analyzes resumes against job descriptions to provide compatibility scores and generate tailored interview questions.

Project Structure

scout-sense/
├── src/
│   ├── models/
│   │   ├── __init__.py
│   │   └── schemas.py          # Data models and schemas
│   ├── services/
│   │   ├── __init__.py
│   │   ├── document_parser.py  # PDF and DOCX parsing
│   │   ├── job_scraper.py     # Job description scraping
│   │   └── ai_analyzer.py     # AI analysis
│   ├── utils/
│   │   ├── __init__.py
│   │   ├── config.py          # Configuration management
│   │   └── constants.py       # Environment variables and constants
│   ├── __init__.py
│   └── main.py                # Main Streamlit application
├── config.yaml                # Application configuration
├── .env                       # Environment variables
├── .env.example              # Example environment variables
├── .gitignore
├── README.md
└── requirements.txt

✨ Features

• 📄 Upload resume in PDF or DOCX format
• 🔗 Provide job description URL for analysis
• 🤖 Multiple AI model options:
  • OpenAI Models:
    • GPT-4o (gpt-4o)
    • GPT-4o Mini (gpt-4o-mini)
  • Google Models:
    • Gemini 1.5 Pro
    • Gemini 1.5 Flash
• ⚙️ Interactive configuration:
  • 🔑 Real-time API key management
  • 🎛️ Adjustable model parameters:
    • 🌡️ Temperature control
    • 📝 Token limits
    • ⚡ Model-specific settings
• 🎯 AI-powered resume analysis and scoring
• 💡 Detailed feedback on candidate's fit for the role
• ❓ Automatic generation of interview questions (for scores ≥ 7/10)
• 🎯 Questions categorized by difficulty:
  • 🟢 Easy
  • 🟡 Intermediate
  • 🟠 Difficult
  • 🔴 Extremely Difficult

⚙️ Configuration

The application offers multiple ways to configure its behavior:

🔑 API Keys

You can provide API keys in Streamlit UI:

• Input keys directly in the sidebar
• Keys are stored securely in the session

🎛️ Model Settings

Configure model parameters through:

1. config.yaml file:

   models:
     gpt4o:
       temperature: 0.7
       max_tokens: 2048
     gemini_pro:
       temperature: 0.7
       max_output_tokens: 2048

2. Interactive UI:
   • Adjust temperature
   • Set token limits
   • Update settings in real-time

📋 Prerequisites

• Python 3.10+
• OpenAI API key (for GPT-4o, GPT-4o-mini)
• Google API key (for Gemini Pro, Gemini 1.5 Pro, Gemini 1.5 Flash)
• Conda (for environment management)

🚀 Installation

1. Clone the repository:

git clone https://github.com/di37/scoutsense-recruit
cd scoutsense-recruit

2. Create and activate a conda environment:

conda create -n scout-sense python=3.10
conda activate scout-sense

3. Install the required packages:

pip install -r requirements.txt

4. Create a .env file:

cp .env.example .env

5. Add your API keys to the .env file or provide them through the UI.

🏃‍♂️ Running the Application

1. Activate the conda environment if not already activated:

conda activate scout-sense

2. Start the Streamlit server:

python main.py

3. Open your web browser and navigate to:

http://localhost:8501

📖 Usage

1. 🔑 Configure API Keys:

   • Enter your OpenAI and/or Google API keys in the sidebar
   • Keys are required only for the models you plan to use

2. 🤖 Select and Configure Model:

   • Choose your preferred AI model from the dropdown
   • Adjust model parameters if desired:
     • Temperature (controls creativity)
     • Token limits (controls response length)

3. 📊 Analyze Resume:

   • Enter the job description URL
   • Upload a resume (PDF or DOCX)
   • Click "Analyze Resume"

4. 📈 View Results:

   • Compatibility score (0-10)
   • Detailed feedback
   • Interview questions (if score ≥ 7)

📸 Screenshots

🧩 Components

📊 Models

• schemas.py: Defines data models for job descriptions and analysis results

🛠️ Services

• document_parser.py: Handles PDF and DOCX file parsing
• job_scraper.py: Scrapes and processes job descriptions from URLs
• ai_analyzer.py: Manages AI model interactions and analysis

⚙️ Utils

• config.py: Configuration management and validation
• constants.py: Environment variables and system constants

🛠️ Technologies Used

• Streamlit
• OpenAI GPT-4o Series
• Google Gemini Series
• PyPDF2
• python-docx
• Beautiful Soup 4
• Pydantic
• PyYAML

📝 Notes

• The application requires an active internet connection to:
  • 🌐 Scrape job descriptions
  • 🤖 Communicate with AI APIs
• 📊 Large resumes or complex job descriptions may take longer to process
• 📈 The quality of the analysis depends on:
  • Clarity of resume and job description
  • Selected model and its configuration
  • Quality of extracted text from documents

---

Made with ❤️ using Streamlit • Powered by Scout Sense 🎯