Analyze. Synthesize. Strategize.
A powerful RAG (Retrieval-Augmented Generation) platform
fueled by Gemini 2.5 Flash and LangGraph.
Transforming static documents, PDFs, and web links into interactive, actionable research notebooks.
|
|
- 🤖 Agentic Reasoning Loop: Powered by LangGraph, allowing the AI to "think" through complex queries rather than just retrieving text.
- 🧠 Enterprise RAG Pipeline: Orchestrated by LangChain for precise document chunking, citation, and hallucination reduction.
- 🔑 Bring Your Own Key (BYOK): A secure, user-centric architecture that requires users to provide their own Gemini API key (AES-encrypted in the database) to use the app, ensuring zero LLM costs for the host.
- 🛡️ Production-Ready Guardrails: Hardened with Upstash Redis rate-limiting (req/min & req/day) and Supabase row-level security quotas (max notebooks/files) to protect the backend from abuse.
- 🔍 Deep Semantic Search: Utilizes Supabase pgvector to find hidden connections across massive PDF uploads and web pages.
- 📊 Live Analytics Dashboard: A highly responsive, collapsible React interface featuring real-time state synchronization via custom dispatch events. Fully optimized for both Desktop and Mobile.
Desktop View (Collapsible Sidebar) |
Mobile View |
Immersive Reading Mode |
Mobile Chat Interface |
Context Upload |
Mobile View |
User Preferences |
Mobile Config |
cd backend
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install -r requirements.txt
cd ../frontend
npm install
Create .env files in both backend and frontend directories following the .env.example templates.
Terminal 1 — Backend
cd backend
python -m uvicorn main:app --reloadTerminal 2 — Frontend
cd frontend
npm run dev