New finding just dropped? Find out what it means for your research team in seconds
Cascade is an AI-powered research analysis platform that helps researchers validate claims by detecting contradictions, mapping citation cascades, and generating strategic insights. Upload a research paper or paste a claim, and the system will:
- Detect Contradictions - Find papers that challenge your findings
- Map Citation Cascades - Trace how ideas spread through academia
- Generate Strategic Insights - Provide actionable research directions
Powered by NVIDIA NeMo and real-time Perplexity API integration.
Research validation is time-consuming and often incomplete. Researchers need to:
- Manually search for contradictory findings
- Trace citation networks to understand impact
- Synthesize insights across multiple papers
- Stay current with rapidly evolving fields
Traditional literature review methods are slow, subjective, and miss important connections.
Cascade automates research validation through a three-stage AI agent system:
- Analyzes your research claim using Perplexity API
- Identifies papers that directly contradict your findings
- Extracts key excerpts and reasoning
- For each contradictory paper, finds papers that cite it
- Maps the impact chain through academia
- Identifies research trends and directions
- Combines contradiction and citation data
- Generates actionable research recommendations
- Highlights gaps and opportunities
┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
│ Frontend │ │ Backend │ │ AI Agents │
│ (React/TS) │◄──►│ (FastAPI) │◄──►│ (Python) │
└─────────────────┘ └─────────────────┘ └─────────────────┘
│ │ │
│ │ │
┌─────────┐ ┌─────────┐ ┌─────────┐
│ Vite │ │ Uvicorn │ │Perplexity│
│ Dev │ │ Server │ │ API │
└─────────┘ └─────────┘ └─────────┘
- Frontend: React 18, TypeScript, Tailwind CSS, Framer Motion
- Backend: FastAPI, Python 3.8+, Uvicorn
- AI: Perplexity API, NVIDIA NeMo integration
- Build: Vite, npm
- Python 3.8+ with virtual environment
- Node.js 18+
- Perplexity API Key (get here)
-
Clone and setup:
git clone <repository-url> cd research-demo pip install -r requirements.txt
-
Configure API:
cp env.example .env # Add your Perplexity API key to .env -
Launch:
# Terminal 1: Start backend python ui/backend_server.py # Terminal 2: Start frontend cd ui npm install npm run dev
-
Access: Open
http://localhost:3000
POST /extract_text- Extract text from PDF filesPOST /detect_contradictions- Find contradictory research papersPOST /propagate_citations- Map citation cascadesPOST /generate_synthesis- Generate research strategy
- Instant results via Perplexity API
- Progressive UI with step-by-step updates
- Live web search integration
- Modular architecture with specialized agents
- NVIDIA NeMo orchestration
- Citation intelligence and network analysis
- PDF upload with drag-and-drop
- Text input for direct claims
- Responsive design for all devices
- Validate breakthrough claims before publication
- Identify gaps in existing research
- Understand citation impact of findings
- Collaborative literature review
- Automated contradiction detection
- Strategic research planning
- Quality assurance for research claims
- Resource optimization for high-impact areas
- Knowledge synthesis across domains
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
This project is licensed under the MIT License - see the LICENSE file for details.
- NVIDIA NeMo - AI orchestration capabilities
- Perplexity AI - Real-time research search and analysis
- Open Source Community - Tools and libraries
Built by researchers, for researchers