ScholarFlow
AI Research Paper Analysis Agent
A full-stack research agent to upload, fetch, summarize, and chat with research papers — then organize whole libraries, assess rigor & bias, and synthesize findings. Built with React + Vite, FastAPI + Mangum, PostgreSQL, Google Gemini, and Vercel Blob, deployed on Vercel.
Main Features
Upload a paper, then read smarter — summarize, question, compare, and synthesize at scale.
Upload & Fetch
Add papers by file upload or fetch by arXiv ID, DOI, or link — with bulk-select and bulk-delete in the library.
AI Summarization
Gemini extracts summary, key elements, keywords, readability, and complexity for every paper automatically.
RAG Q&A Chat
Ask questions of a paper and get grounded answers, with clickable follow-up question suggestions.
Collections & Status
Organize papers into collections with reading status (not started → completed) and progress tracking.
Literature Review
Generate a structured review (Intro / Methodology / Findings / Gaps / Future Work) with tone and length options.
Related Papers
Embedding-centroid cosine similarity ranks related work and shows a match percentage per paper.
Rigor & Bias
Each paper gets rigor, reproducibility, and bias-risk scores with quality flags shown as badges.
Research Questions
Track research questions & hypotheses per collection with active/resolved/abandoned states.
Comparison Matrix
Compare methodologies across papers (sample size, duration, data source, controls) with CSV export.
Analytics Dashboard
Topics, distributions, monthly velocity, and average rigor/reproducibility across your library.
Translate & Export
Translate summaries into many languages; export to Markdown, BibTeX, Notion, or Obsidian.
Notes & Highlights
Attach plain-text notes and highlights (with page numbers) to any paper in your library.
Technology Stack
More Projects
Browse all →Build your AI research tool
From React + FastAPI agents to Gemini-powered RAG and synthesis, we ship intelligent research platforms end to end.