Resume RAG Chatbot

A retrieval-augmented generation (RAG) system that lets recruiters query your resume conversationally instead of scanning bullet points.

Live Demo: https://saniresume-rag.streamlit.app/

Features

• Conversational Q&A - Ask questions like "What experience do you have with distributed systems?"
• Semantic Search - BGE-base-en-v1.5 embeddings for accurate retrieval
• Fast Vector Index - FAISS for efficient similarity search
• LLM Powered - Llama 3.3 70B via Groq API
• Source Attribution - Shows exactly which resume sections informed each answer
• Guardrails - Ensures answers stay grounded in your actual experience

Tech Stack

Component	Technology
Frontend	Streamlit
Embeddings	BGE-base-en-v1.5 (Sentence Transformers)
Vector Store	FAISS
LLM	Llama 3.3 70B via Groq
Language	Python 3.12+

Quick Start

1. Install Dependencies

pip install -r requirements.txt

2. Set Environment Variables

export GROQ_API_KEY=your_groq_api_key

Get a free Groq API key at https://console.groq.com/keys

3. Add Your Resume

Edit resume.md with your actual resume content, then pre-compute the embeddings:

python precompute_index.py

4. Run the App

streamlit run streamlit_app.py

How It Works

User Query
    ↓
┌─────────────────────────────────────────┐
│         Embed Query (BGE-base)          │
└─────────────────────────────────────────┘
    ↓
┌─────────────────────────────────────────┐
│     FAISS Vector Search (Top-K)         │
└─────────────────────────────────────────┘
    ↓
┌─────────────────────────────────────────┐
│   Context Injection + Guardrails        │
└─────────────────────────────────────────┘
    ↓
┌─────────────────────────────────────────┐
│      Llama 3 Inference (Groq)           │
└─────────────────────────────────────────┘
    ↓
   Answer

File Structure

resume-rag/
├── streamlit_app.py      # Main Streamlit application
├── rag_engine.py         # RAG pipeline (embeddings, FAISS, Groq)
├── precompute_index.py   # Pre-compute embeddings for faster startup
├── requirements.txt      # Python dependencies
├── resume.md             # Your resume content
├── README.md             # This file
└── .gitignore            # Git ignore rules

Example Questions

• "What experience do you have with distributed systems?"
• "Have you built production RAG pipelines?"
• "How have you improved inference throughput?"
• "What's your experience with vector databases?"
• "Tell me about your ML projects"

Customization

Change the Embedding Model

Edit rag_engine.py:

self.embed_model = SentenceTransformer("BAAI/bge-large-en-v1.5")  # Larger model

Change the LLM

Edit rag_engine.py:

self.llm_model = "llama-3.1-8b-instant"  # Faster, smaller model

Adjust Chunk Size

Edit precompute_index.py:

chunks = chunk_text(full_text, chunk_size=300, overlap=50)