Textbook becomes mind map, before you even finish reading.
BrainLattice maps the lattice structure of knowledge onto your brain's neural network. Learning designed for how the mind actually works.
Traditional learning forces linear consumption of non-linear information. BrainLattice flips this paradigm by transforming static PDF textbooks into interactive, explorable knowledge graphs. It uses LLMs to extract core concepts, identify relationships, and generate high-fidelity study materials, allowing users to understand the "system" of a subject before memorizing the details.
- Obsidian-like Knowledge Graph - Spatially explore concepts.
Cmd + Shift + Clickto drill down into formulas, theorems, and prerequisites. - Contextual Synthesis - Auto-generates study guides and entity summaries from raw text.
- Audio Briefs - Converts synthesized notes into podcast-style audio via Fish Audio.
- Universal Input - Works with any text-heavy PDF: textbooks, research papers, lecture notes.
- FastAPI + Next.js
- Google Gemini 2.5 Flash - Concept extraction & graph topology.
- Grok 4 (OpenRouter) - Summarization & synthesis.
- Fish Audio - Low-latency and cost-effective TTS.
- React Force Graph - WebGL-powered visualization.
- Firebase - Real-time data & auth.
- Python 3.10+
- Node.js 18+
- Fish Audio API Key
- Google Gemini API Key
- OpenRouter API Key
- Firebase Service Account
-
Backend Setup
cd backend pip install -r requirements.txt uvicorn main:app --reloadServer runs at
http://localhost:8000 -
Frontend Setup
cd frontend npm install npm run devApp runs at
http://localhost:3000
GEMINI_API_KEY=your_gemini_key
OPENROUTER_API_KEY=your_openrouter_key
FISH_AUDIO_API_KEY=your_fish_audio_key
FIREBASE_SERVICE_ACCOUNT_PATH=secrets/firebase_private.json- Ingestion: PyPDF extracts raw text from uploaded documents.
- Synthesis: Grok 4 (via OpenRouter) identifies sequential concepts.
- Mapping: Gemini 2.5 constructs the hierarchical knowledge graph.
- Generation: The system produces study guides and audio scripts.
- Audio: Scripts are streamed to Fish Audio for TTS generation.