AI Resume Analyzer

This project now includes a standalone app that ranks resumes from a local folder or uploaded files using sentence-transformers embeddings. Optional features include reranking with a free Hugging Face cross-encoder and prompt refinement via local Ollama (Llama3).

1) Install Dependencies

pip install -r requirements.txt
# Optional for reranking (uncomment in requirements.txt or install directly):
# pip install transformers torch

2) Launch Gradio

python gradio_app.py

Open the URL shown in the terminal (typically http://127.0.0.1:7860).

3) Use the UI

• Modes:
  • Folder Mode: provide path to a directory containing PDF/DOCX resumes.
  • Upload Mode: upload one or more PDF/DOCX resumes directly.
• Job Description / Query: text describing the role or requirements.
• Top K: number of top matches to display.
• Min Score: filter out low-relevance results.
• Keywords: comma-separated terms to boost (each unique hit adds to score).
• Keyword Boost: weight per unique keyword found in a resume.
• Use HF Reranker: enable cross-encoder reranking (BAAI/bge-reranker-base).
• Refine via Ollama: if ollama serve is running with llama3.2 pulled, query refinement runs locally.
• Export CSV: download current ranking as CSV.
• Preview: view metadata and excerpt for a specific rank.

Notes:

• The embedding model is all-MiniLM-L6-v2 (free on Hugging Face), fast and CPU-friendly.
• Reranking uses BAAI/bge-reranker-base if installed; otherwise the app still works with embeddings only.
• PDF parsing uses PyMuPDF; DOCX parsing uses docx2txt.

---

JSON API: Rank Folder (Django)

Endpoint: POST /rank-folder/

Body (JSON):

{
  "folder_path": "/absolute/path/to/resumes",
  "query": "Senior Python developer with Django",
  "top_k": 10,
  "keywords": "python,django,rest,aws",
  "min_score": 0.2,
  "keyword_boost": 0.05,
  "use_ollama": false,
  "use_reranker": false
}

Response:

{
  "indexed": 12,
  "results": [
    {"rank": 1, "name": "...", "path": "...", "score": 0.87, "base_score": 0.82, "keyword_hits": 3, "excerpt": "..."},
    ...
  ]
}

---

Instructions for Configuration

1. Activate Your Conda Environment

If you already have a Conda environment, activate it:

conda activate shadril238  # Replace with your actual environment name

If you haven’t created one, do:

conda create --name shadril238 python=3.10 -y
conda activate shadril238

---

2. Install Project Dependencies

If your project has a requirements.txt file, install dependencies:

pip install -r requirements.txt

Otherwise, manually install Django and DRF:

pip install django djangorestframework

---

3. Set Up Environment Variables

If your project uses .env files, install python-dotenv:

pip install python-dotenv

Then create a .env file if it doesn’t exist:

touch .env

Add necessary environment variables inside .env:

DJANGO_SECRET_KEY=your_secret_key
DEBUG=True
DATABASE_URL=your_database_url

---

4. Apply Migrations

Run the following command to apply database migrations:

python manage.py migrate

---

5. Create a Superuser (Optional)

If your project has Django Admin, create an admin user:

python manage.py createsuperuser

Follow the prompts to set up a username and password.

---

6. Run the Django Server

Start the development server:

python manage.py runserver

By default, it runs at http://127.0.0.1:8000/.

---

7. Open the Project in Your Browser

• API endpoints (if using DRF): http://127.0.0.1:8000/api/
• Django Admin: http://127.0.0.1:8000/admin/

---

8. (Optional) Debugging Issues

• Check if the correct Conda environment is activated:

  conda info --envs

• If manage.py doesn’t run, check installed packages:

  pip list

Now your Django Conda project is running successfully! 🚀

---