StudyRadar

• 

  StudyRadar

• 

  Dashboard

• 

StudyRadar is an AI-powered research companion that helps students and researchers compare multiple research papers simultaneously. Instead of summarizing a single document, it analyzes 3–4 PDFs together and generates a structured synthesis matrix that highlights common findings, conflicting viewpoints, supporting evidence, and key statistical insights, making literature reviews significantly faster and more organized.

How it was built

Multi-Document RAG (LangChain): Implemented an ensemble retrieval pipeline to extract relevant context across multiple research papers simultaneously. Backend (FastAPI + Groq): Built a FastAPI service that processes uploaded PDFs, retrieves relevant chunks, and uses Groq's LLM to generate a structured JSON output conforming to a predefined comparison matrix schema. Interactive Frontend: Designed a clean dashboard with a color-coded comparison table that visually highlights agreements, contradictions, and important findings for easy analysis.

This project was inspired by the time-consuming process of manually comparing research papers during literature reviews. The biggest challenge was ensuring accurate cross-document retrieval while maintaining a consistent and structured output format across papers with different writing styles and layouts.

Built With

• fastapi
• groq
• langchain