AI Resume Analyzer

An intelligent, full-stack application that evaluates a candidate's resume against a specific job description. This tool uses a hybrid approach, combining mathematical semantic similarity (ATS scoring) with generative AI to provide actionable, qualitative feedback. Features

• Dual-Engine Analysis:
  • ATS Similarity Score: Uses BERT embeddings (all-mpnet-base-v2) to mathematically calculate the cosine similarity between the resume and the job description.
  • Qualitative AI Feedback: Integrates with the Groq API (running Llama-3.3-70b-versatile) to act as an AI recruiter, scoring individual job requirements and providing emoji-coded feedback.
• Separation of Concerns: Built with a modern architecture featuring a high-performance FastAPI backend and an interactive Streamlit frontend.
• Efficient Processing: The heavy machine learning model (BERT) is cached in memory on server startup to ensure lightning-fast subsequent requests.
• Exportable Reports: Users can download their personalized AI analysis directly as a text file.

Tech Stack

• Backend: FastAPI, Uvicorn, Python
• Frontend: Streamlit, Requests
• Machine Learning & NLP: Sentence-Transformers, scikit-learn
• LLM Integration: Groq API (Llama 3)
• Utilities: PDFMiner.six (PDF text extraction), python-dotenv

Project Structure

ai-resume-analyzer/
│
├── backend.py            # FastAPI server, ML models, and Groq API logic
├── frontend.py           # Streamlit user interface
├── requirements.txt      # Python dependencies
├── .env                  # Environment variables (API keys) - Ignored by Git
└── .gitignore            # Git ignore rules

## ⚙️ Installation & Setup

### 1. Clone the Repository

```bash
git clone https://github.com/YourUsername/ai-resume-analyzer.git
cd ai-resume-analyzer

2. Install Dependencies

pip install -r requirements.txt

3. Configure API Key

1. Create an account on Groq and generate an API key.
2. Create a .env file in the root directory of the project.
3. Add the following line to the .env file:

GROQ_API_KEY=gsk_your_actual_api_key_here

---

Running the Application

This project uses a decoupled architecture, so the backend and frontend must be run in two separate terminal windows.

Start the Backend

Open a terminal inside the project folder and run:

uvicorn backend:app --reload

Wait until you see:

Application startup complete

---

Start the Frontend

Open a second terminal in the same project folder and run:

streamlit run frontend.py

The application interface will automatically open in your default web browser.

---

How to Use

1. Upload Resume Upload your resume in PDF format.

2. Paste Job Description Enter the target job description into the provided text area.

3. Analyze Resume Click the Analyze button to generate:

   • ATS compatibility score
   • Skill match analysis
   • AI-generated feedback and suggestions

4. Download Report Save the generated analysis report locally using the Download Report button.