🌊 SubsurfAI — Physics-Aware Subsurface Interpretation Platform

Author: Abdulrahman Al-Fakih · KFUPM / Geoscience AI Lab
Live demo:

SubsurfAI is an open-source, physics-constrained well log interpretation platform for the Dutch North Sea. It combines classical petrophysical equations (Archie's law, Wyllie time-average, NPHI-RHOB crossplot) with AI-powered analysis using Claude.

---

What it does

• Upload any LAS well log → instant petrophysical interpretation
• 6 bundled Dutch North Sea wells (F-block, NLOG) for instant demo
• Physics-constrained lithofacies (Shale / Sandstone / Limestone / Coal)
• Reservoir quality — porosity (φ_eff), water saturation (Sw), uncertainty bands
• AI Chat — ask questions about your well in plain English, get expert answers
• Multi-well comparison — compare up to 4 wells side by side
• Export — download interpreted CSV or LAS with new curves

---

Foundation model

GlobalWellFM — trained on 163,000 depth samples from 3 countries:
🤗 Aljaser2030/GlobalWellFM · F1 = 0.9494

---

Run locally

git clone https://github.com/YOUR_USERNAME/SubsurfAI
cd SubsurfAI
pip install -r requirements.txt
streamlit run app.py

Set your Anthropic API key in .env:

ANTHROPIC_API_KEY=sk-ant-...

---

Citation

Al-Fakih, A. (2026). SubsurfAI: A Physics-Constrained AI Platform for Well Log
Interpretation in the Dutch North Sea. KFUPM / Geoscience AI Lab.

---

License

MIT — free to use, modify, and distribute with attribution.