INTERACTIVE PROTOTYPE: This is a static frontend UI mock-up of GeoBot. It mimics the design of the real app but is not connected to the live backend AI.
🌍 GeoBot - Geology Expert Chatbot
Back to Showcase
Fact-checking & Expert Q&A powered by 111k+ peer-reviewed research papers | orebit.id
Knowledge Base: ● ONLINE | 111,000+ Peer-Reviewed Papers

đŸ’Ŧ Expert Chat Interface

đŸŽ¯ PINN-Geostatistics Intelligence:
✓ PINN Expert: Specialized in Physics-Informed Neural Networks
✓ Fact-checking against 111k+ peer-reviewed papers (RAG)
✓ PhD-Level Analysis on geostatistics & resource estimation
✓ Gemini AI powered reasoning engine
🤖 GeoBot (PINN v2.3 Edition)

👋 Welcome! I specialize in Physics-Informed Neural Networks for Geostatistics.

đŸŽ¯ My Capabilities:
📚 Literature Review: 111k+ peer-reviewed papers (RAG-powered)
đŸ”Ŧ PINN Theory: Physics Loss functions, domain awareness, normalization
⚖ī¸ Challenger Mode: Benchmarking PINN vs Kriging (RMSE vs Variance)
✅ Production Validation: Swath plots, drift analysis, JORC compliance

🔗 Live Integration: GeoDataViz Pro v2.3 (app.orebit.id)

Ask about PINN implementation, benchmarking strategies, or request literature citations! (Note: This is a UI demo)
📝 Expert Knowledge Templates:

ℹī¸ About GeoBot

📚 RAG Architecture:
Embeddings: Ollama (Local/Fast)
Vector Store: DuckDB (Production)
LLM: Gemini 1.5 Flash
Knowledge: 111k+ peer-reviewed papers

Sources: Elsevier â€ĸ Springer â€ĸ AGU
đŸŽ¯ Research Topics:
Mineral deposits, geochemistry, structural geology, exploration methods, geostatistics, resource estimation, geophysics
🧠 Technology:
RAG (Retrieval-Augmented Generation) with DuckDB vector store & Gemini AI

⚙ī¸ System Info

Version: 9.0.0 Production (PINN v2.3)

Backend: n8n RAG (Gemini + DuckDB) (Simulated)

PINN Engine: TensorFlow (GeoDataViz Pro)

Status: ● Online

Session: s_20260220_demo