AI-Based Resume Screener

This repository hosts an AI-based resume screening system designed to scale and handle large volumes of resumes. Built using a Retrieval-Augmented Generation (RAG) pipeline, this system enables efficient resume ingestion and querying, utilizing cutting-edge LLMs and a simple-to-use Flask-based API interface.

Table of Contents

• Features
• Architecture
• Setup
• Usage
• File Structure
• License

---

Features

• RAG Pipeline: Retrieval-Augmented Generation (RAG) setup using Langchain and ChromaDB.
• LLM Integration: Supports Gemini-1.5-Flash LLM API and an alternative Local Llama 3.2:3b model.
• Scalability: Efficiently processes and scales with over 500 resumes.
• Flask API: Provides backend APIs to interface the RAG pipeline and LLM with the frontend.

Architecture

The AI-Based Resume Screener comprises the following main components:

1. Ingestion and Storage:

   • Resumes are ingested in a format that allows for easy querying.
   • ChromaDB stores resume embeddings for fast retrieval.

2. Retrieval-Augmented Generation Pipeline:

   • Built using Langchain and ChromaDB.
   • Ensures the system finds relevant information in resumes quickly and accurately.

3. LLM Integration:

   • Primary: Gemini-1.5-Flash LLM for high-quality language understanding.
   • Alternative: Local Llama 3.2:3b for deployments where Gemini-1.5 access is limited.

4. Backend with Flask:

   • Flask-based backend APIs allow seamless communication between the frontend and the RAG pipeline.

---

Setup

To set up this project locally, follow these steps:

Prerequisites

• Python 3.8+
• Docker (optional for Local Llama)
• API keys:
  • Gemini-1.5-Flash LLM API key (required for LLM integration)
• Packages:
  • pip install -r requirements.txt (see instructions below)

Installation

1. Clone the Repository:

   git clone https://github.com/yourusername/ai-based-resume-screener.git
   cd ai-based-resume-screener

2. Install Required Packages:

   pip install -r requirements.txt

3. Set Up Environment Variables: Create a .env file to store your API keys and configuration. Example:

   GEMINI_API_KEY=your_gemini_api_key_here

4. Run Local Llama (Alternative LLM): If you choose to use Local Llama, ensure Docker is installed, and run:

   docker pull local_llama_image
   docker run -d -p 5001:5001 local_llama_image

5. Run the Flask App:

   flask run

---

Usage

Once the server is up and running, you can interact with the backend APIs for querying resumes. You can also connect a frontend application to visualize and query the resume data.

API Endpoints

• /api/ingest:

  • Method: POST
  • Description: Ingests a new resume into the system.

• /api/query:

  • Method: POST
  • Description: Queries the resumes based on the input question using the RAG pipeline.

Sample Requests

Ingest Resume:

curl -X POST -H "Content-Type: application/json" -d '{"resume_text": "John Doe, experienced data scientist..."}' http://localhost:5000/api/ingest

Query Resume:

curl -X POST -H "Content-Type: application/json" -d '{"question": "What is the experience of the candidate in data science?"}' http://localhost:5000/api/query

---

File Structure

• app/: Contains the main Flask application code.
• rag_pipeline/: Houses the RAG pipeline with Langchain and ChromaDB integration.
• models/: Directory for LLM model wrappers (Gemini and Local Llama).
• data/: Directory for storing resumes (as needed).
• requirements.txt: List of required packages.
• .env: Environment file to store API keys and secrets (not included in version control).

---

License

This project is licensed under the MIT License.