About

|

ScholaRAG democratizes access to systematic literature review automation.

We believe researchers should spend more time analyzing findings and less time on mechanical tasks. By combining PRISMA 2020 guidelines with modern RAG technology, we help academics conduct rigorous reviews in weeks instead of months.

Core Capabilities

Complete Retrieval

Fetch all available papers from multiple academic databases with comprehensive coverage and direct PDF access.

20K+ papers · 5 databases · 100% coverage

AI-PRISMA Screening

Multi-dimensional paper evaluation using LLMs with transparent criteria, confidence scoring, and human validation.

PICO framework · Cohen's κ · 10-20% precision

Custom RAG Systems

Build semantic search over your papers using vector databases for instant literature queries with citations.

ChromaDB · FAISS · Contextual retrieval

Conversational Workflow

Step-by-step guidance through Claude Code in VS Code, making systematic reviews accessible without coding.

7 stages · 4-5 hours · Interactive prompts

Technology

AI Models
  • Claude Sonnet 4.5
  • Claude Haiku 4.5
  • GPT-5-Codex
Databases
  • Semantic Scholar
  • OpenAlex
  • arXiv
  • Scopus
Vector Stores
  • ChromaDB
  • FAISS
  • Qdrant
Framework
  • Python 3.9+
  • LangChain
  • VS Code

State-of-the-Art AI Coding Models

ScholaRAG leverages the latest AI coding models optimized for research automation, keeping operational costs under $20-40 per project for typical systematic reviews.

Claude Sonnet 4.5

Released October 2025 - currently the most effective coding model for research automation. Achieves state-of-the-art performance on SWE-bench for complex workflow generation.

Best for automation · Advanced reasoning

GPT-5-Codex

Advanced code generation with superior reasoning for complex research workflows. Excellent for systematic review pipeline design and execution.

Advanced codegen · Research-optimized

Research Impact

PhD Students

2-3 weeks vs 6-8 weeks

Dissertation literature reviews, qualifying exams, comprehensive analysis

Academic Researchers

67-75% time savings

Meta-analysis preparation, grant proposals, PRISMA 2020 reviews

Professors & Faculty

Never forget citations

Course updates, research synthesis, student mentoring

Research Librarians

Scalable support

Systematic review consulting, evidence-based practice workshops

Open Source & Free

MIT Licensed · Community Driven · Fully Transparent

Use for academic or commercial research

Customize workflows for specific domains

Contribute improvements and templates

Run locally with full data control

View on GitHub

Contact

Project Maintainer

Hosung You

PhD Candidate

College of Education

The Pennsylvania State University

Get in Touch

hfy5138@psu.edu

Bug reports: GitHub Issues

Questions: GitHub Discussions

Collaboration: Email

Citation

If you use ScholaRAG in your research:

@software{scholarag2025,
  author = {You, Hosung},
  title = {ScholaRAG: AI-Powered Systematic
           Literature Review Automation},
  year = {2025},
  url = {https://github.com/HosungYou/ScholaRAG},
  note = {Open-source PRISMA-compliant RAG system}
}

Ready to transform your literature review?

Get Started