Skip to content

danielxies/scholarflow

Repository files navigation

ScholarFlow

AI-powered academic research platform that automates the tedious parts of the research workflow — literature discovery, experiment replication, and paper writing — in a single browser-based IDE.

What It Does

ScholarFlow takes a research idea and turns it into a complete paper draft by automating the grunt work:

  1. Literature Search — Finds and ranks relevant papers via OpenAlex, adds them to your library
  2. Experiment Replication — Runs baseline experiments on cloud GPUs (Modal) to validate prior work
  3. Novel Experiments — Designs and executes new experiments based on your hypothesis
  4. Paper Writing — AI drafts sections with proper citations, renders LaTeX to PDF in real-time

Tech Stack

Category Technologies
Frontend Next.js 16, React 19, TypeScript, Tailwind CSS 4
Editor CodeMirror 6, LaTeX preview via latex.js
Backend SQLite (local-first), Inngest (background jobs)
AI Claude Sonnet 4 (Anthropic API)
Compute Modal (GPU experiment execution)
Search OpenAlex API (academic paper discovery)
UI shadcn/ui, Radix UI, Streamdown (markdown rendering)

Features

  • LaTeX Editor with live PDF preview and syntax highlighting
  • AI Chat with streaming responses and tool call visualization
  • Paper Library with relevance scoring and enrichment
  • Experiment Workspace with hypothesis tracking, findings, and research logs
  • Guided Demo Mode for presentations and walkthroughs

Getting Started

# Install dependencies
npm install

# Set up environment variables
cp .env.example .env.local
# Fill in: ANTHROPIC_API_KEY, OPENALEX_API_KEY

# Run the dev server
npm run dev

# In another terminal, start Inngest
npx inngest-cli@latest dev

Environment Variables

Variable Required Description
ANTHROPIC_API_KEY Yes Claude API key
OPENALEX_API_KEY No OpenAlex API key (higher rate limits)
REPRODUCTION_RUNNER_BACKEND No Set to modal for GPU experiments
REPRODUCTION_MODAL_SUBMIT_URL No Modal worker endpoint

Deployment

The app deploys as a single service on Railway:

railway login
railway init
railway up

SQLite runs on Railway's persistent disk — no external database needed.

Team

Built for the hackathon by the ScholarFlow team.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors