AxorynAI

Design and Development of an Intelligent Research Assistant using Full-Stack Technologies, Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs)

AxorynAI is an intelligent AI-powered research assistant designed to help researchers, students, and professionals efficiently organize, analyze, retrieve, and interact with knowledge from large collections of documents through natural language conversations.

The system combines modern full-stack technologies with Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs) to provide context-aware, grounded, and citation-supported responses while minimizing hallucinations commonly observed in standalone language models.

This project was developed as part of the Summer Internship 2026 at National Institute of Technology Jamshedpur (NIT Jamshedpur) under the guidance of Dr. Dilip Kumar.

---

Features

Authentication & Security

• JWT Authentication with Access and Refresh Tokens
• Secure Password Hashing
• Google OAuth 2.0 Login
• Protected API Routes
• Session Management
• Role-Based User Access
• Token Refresh Mechanism

Workspace Management

• Create and Manage Multiple Research Workspaces
• Workspace Isolation for Independent Research Projects
• Workspace-Specific Context Management
• Persistent User Sessions

Intelligent Document Processing

• PDF Document Upload and Processing
• DOCX Document Upload and Processing
• Automatic Text Extraction and Cleaning
• Chunking and Semantic Segmentation
• Metadata Preservation

Retrieval-Augmented Generation (RAG)

• Semantic Vector Search
• Context Retrieval Pipeline
• Similarity-Based Document Ranking
• Citation-Aware Responses
• Multi-Document Knowledge Retrieval
• Hallucination Reduction through Grounded Context

AI Capabilities

• Natural Language Question Answering
• Research Summarization
• Context-Aware Conversations
• Multi-Turn Dialogue Support
• Research Knowledge Exploration
• Intelligent Information Retrieval

User Experience

• Modern Responsive Interface
• Workspace Dashboard
• Research Chat Interface
• Real-Time Interaction
• Document Management System
• Profile Management

---

Technology Stack

Frontend

• React.js
• Vite
• Tailwind CSS
• React Router
• Axios

Backend

• FastAPI
• SQLAlchemy
• PostgreSQL
• JWT Authentication
• Google OAuth

Artificial Intelligence

• Retrieval-Augmented Generation (RAG)
• Large Language Models (LLMs)
• Grok API
• Semantic Embeddings
• Vector Similarity Search

Infrastructure & DevOps

• Docker
• Git
• GitHub
• REST APIs

---

System Architecture

User Query
    ↓
Frontend (React + Vite)
    ↓
FastAPI Backend
    ↓
Authentication & Workspace Validation
    ↓
RAG Retrieval Pipeline
    ↓
Relevant Context Retrieval
    ↓
LLM Processing (Grok API)
    ↓
Grounded Response Generation
    ↓
Response Returned to User

---

Project Structure

AI-Research-Assistant/
│
├── app/
│   ├── core/
│   ├── models/
│   ├── routers/
│   ├── schemas/
│   ├── services/
│   ├── utils/
│   ├── prompts/
│   └── main.py
│
├── frontend/
│   ├── src/
│   │   ├── pages/
│   │   ├── components/
│   │   ├── services/
│   │   ├── context/
│   │   └── assets/
│   │
│   └── public/
│
├── uploads/
├── vector_store/
├── requirements.txt
├── docker-compose.yml
└── README.md

---

Installation

Clone Repository

git clone https://github.com/ImRAryan/AI-Research-Assistant.git
cd AI-Research-Assistant

Backend Setup

python -m venv venv
source venv/bin/activate

For Windows:

venv\Scripts\activate

Install dependencies:

pip install -r requirements.txt

Start the backend server:

uvicorn app.main:app --reload

---

Frontend Setup

cd frontend
npm install
npm run dev

---

Environment Variables

Create a .env file in the backend directory.

DATABASE_URL=
SECRET_KEY=
ALGORITHM=
ACCESS_TOKEN_EXPIRE_MINUTES=

GOOGLE_CLIENT_ID=
GOOGLE_CLIENT_SECRET=

GROK_API_KEY=

---

API Documentation

Interactive API documentation is available through Swagger UI.

Backend API:

https://imraryan-research-assistant.hf.space/docs

Redoc Documentation:

https://imraryan-research-assistant.hf.space/redoc

---

Deployment

Frontend

https://ai-research-assistant-three-wine.vercel.app/

Backend

https://imraryan-research-assistant.hf.space

API Documentation

https://imraryan-research-assistant.hf.space/docs

---

Screenshots

Authentication System

Workspace Management

Research Chat Interface

Document Upload and Processing

API Documentation

---

Research Objectives

• Develop an intelligent document-centric research assistant.
• Integrate Retrieval-Augmented Generation for grounded responses.
• Reduce hallucinations in LLM-generated outputs.
• Enable efficient knowledge discovery from large document collections.
• Provide scalable and modular research infrastructure.

---

Future Enhancements

• Hybrid Retrieval Mechanisms
• Multi-Modal Document Understanding
• Support for Additional Document Formats
• Collaborative Research Workspaces
• Advanced Citation Generation
• Knowledge Graph Integration
• Real-Time Collaboration Features

---

Academic Information

Project Title:

Design and Development of an Intelligent Research Assistant using Full-Stack Technologies, Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs).

Institution:

National Institute of Technology Jamshedpur (NIT Jamshedpur)

Internship:

Summer Internship 2026

Academic Guide:

Dr. Dilip Kumar

---

Author

GitHub: https://github.com/ImRAryan

---

Acknowledgements

I express sincere gratitude to Dr. Dilip Kumar for his guidance, mentorship, and support throughout the development of this project during the Summer Internship 2026 at National Institute of Technology Jamshedpur (NIT Jamshedpur).

---

Star the Repository

If you find this project useful, consider giving the repository a star.