Smart Research Assistant (R-Agents)

A GenAI-powered Smart Research Assistant designed to automate the end-to-end academic research workflow — from literature discovery to LaTeX-ready paper generation — within a single unified platform.

---

🚀 Overview

The exponential growth of academic publications has made manual research workflows time-consuming and error-prone. Smart Research Assistant (R-Agents) addresses this challenge by combining Generative AI, scholarly APIs, and a modular full-stack architecture to assist students, researchers, reviewers, and faculty.

The system automates:

• Literature surveys
• Research gap identification
• Methodology drafting
• Citation validation
• Multilingual translation
• Persona-based summarization
• Interdisciplinary synthesis
• LaTeX report generation

All features operate inside a single, scalable, and ethical research ecosystem.

---

✨ Key Features

• 📚 Literature Survey Generator – Fetches and synthesizes papers from Semantic Scholar, OpenAlex, CrossRef, and Unpaywall
• 🔍 Research Gap Finder – Identifies underexplored areas and limitations
• 🧪 Methodology Builder – Generates structured experimental or procedural sections
• 📖 Citation Validator – DOI-based reference verification with APA/IEEE formatting
• 👤 Persona-Based Summarizer – Tailors summaries for students, reviewers, or faculty
• 🌍 Multilingual Translator – Cross-language paper translation and summarization
• 🔗 Cross-Domain Synthesizer – Connects insights across unrelated research fields
• ⚠️ Contradiction Analyzer – Detects claim–evidence inconsistencies
• 📊 Benchmark Explorer – Suggests datasets, metrics, and evaluation standards
• 🧾 LaTeX Generator – Produces publication-ready .tex files
• 🎙️ Voice/Text I/O – Speech input and TTS output for accessibility

---

🧠 Tech Stack

Frontend

• React
• Tailwind CSS

Backend

• FastAPI (Async, Modular)
• Python 3.10+

Database

• PostgreSQL
• SQLAlchemy ORM

AI / LLM Layer

• OpenAI GPT-4o
• Google Gemini 1.5 Pro
• Switchable LLM orchestration (llm_utils.py)

External APIs

• Semantic Scholar
• OpenAlex
• CrossRef
• Unpaywall

---

🏗️ System Architecture

1️⃣ High-Level Architecture

┌────────────────────┐
│   React Frontend   │
│ (Tailwind UI)      │
└─────────▲──────────┘
          │ REST APIs
┌─────────┴──────────┐
│   FastAPI Backend  │
│ (Async Modules)    │
└─────────▲──────────┘
          │
┌─────────┴──────────┐
│   LLM Orchestrator │◄── GPT-4o / Gemini
│   (llm_utils.py)   │
└─────────▲──────────┘
          │
┌─────────┴──────────┐
│ Scholarly APIs     │
│ (SS, OpenAlex,     │
│  CrossRef, etc.)   │
└────────────────────┘
          │
┌────────────────────┐
│ PostgreSQL Database│
└────────────────────┘

---

2️⃣ Layered Architecture

[ UI Layer ]
React + Tailwind CSS

[ Application Layer ]
FastAPI Routes (Modular Services)

[ AI / LLM Layer ]
Prompt Templates + Model Switching

[ Data Layer ]
PostgreSQL + ORM

[ File Layer ]
PDF Uploads | .tex Exports | Audio

---

3️⃣ Module-Level Architecture

Each research feature is implemented as an independent module:

/api
 ├── literature
 ├── gap_analysis
 ├── methodology
 ├── citation_validator
 ├── translator
 ├── persona_summary
 ├── cross_domain
 ├── contradiction
 ├── latex_generator
 └── benchmarks

This ensures:

• Scalability
• Maintainability
• Parallel execution

---

🗂️ Database Design (ER Overview)

User ──┬── Project ──┬── GeneratedOutputs
       │             └── Citations
       └── Roles

• Secure JWT-based authentication
• Hashed passwords
• Role-based access (Student / Researcher / Reviewer / Faculty)

---

🔄 Data Flow

User Input (Text / PDF / Voice)
        ↓
Preprocessing & Cleaning
        ↓
Scholarly Retrieval APIs
        ↓
LLM Synthesis & Validation
        ↓
Database Storage
        ↓
LaTeX / Text / Audio Output

---

🧪 Non-Functional Properties

• ⚡ Asynchronous processing
• 🔐 Secure API key handling via .env
• 📦 Portable (Local / Cloud / Docker-ready)
• 📈 Horizontally scalable
• ♻️ Easily extensible modules

---

⚖️ Ethical & Legal Compliance

• Uses open-access research only
• No training on user data
• Verifiable citations
• Responsible AI usage

---

🛠️ Installation (Local)

git clone https://github.com/your-username/smart-research-assistant.git
cd smart-research-assistant

pip install -r requirements.txt

# Set environment variables
cp .env.example .env

uvicorn main:app --reload

---

📌 Future Enhancements

• Dockerized deployment
• PDF citation highlighting
• Collaborative research spaces
• Offline LLM support
• Plagiarism similarity checks

---

👥 Team

Research Minds

• Ayan Chattopadhyay
• Arkapravo Mandal
• Atreyee Das
• Deboparna Das

---

⭐ If you like this project

Give it a ⭐ on GitHub and feel free to fork or contribute!