RAG Sample: Resume & Job Description Matching

This project demonstrates a simple Retrieval-Augmented Generation (RAG) pipeline using LangChain, FAISS, and Groq LLM. It matches a candidate's resume with a job description to answer questions such as suitability for a role.

Features

• Loads and processes both a resume (PDF) and a job description (TXT)
• Splits documents into manageable chunks for efficient retrieval
• Embeds documents using HuggingFace sentence transformers
• Stores embeddings in a FAISS vector database
• Uses Groq's Llama3-70B model for question answering over the retrieved context

Project Structure

.
├── app.py
├── .env
├── requirements.in
├── docs/
│   └── franco_resume_en.pdf
├── JD/
│   └── 01_falabella_product_owner.txt
└── README.md

Getting Started

1. Install Dependencies

uv pip install -r requirements.in

2. Set Up Environment Variables

Create a .env file with your API keys (e.g., for Groq).

3. Run the Application

python app.py

How It Works

• The script loads both the resume and job description.
• Documents are split into overlapping chunks.
• Chunks are embedded and indexed in FAISS.
• A question is asked (e.g., "¿puede Franco Cedillo postular al puesto de Product Owner en Falabella?").
• The system retrieves relevant chunks and uses the LLM to generate an answer.

Customization

• Replace docs/franco_resume_en.pdf with your own resume.
• Replace JD/01_falabella_product_owner.txt with any job description.
• Modify the query in app.py to ask different questions.

Requirements

See requirements.in for the full list.

License

MIT License

---

Inspired by modern RAG pipelines for HR and recruitment use cases.