Robin Thibaut
Head of Agentic Systems R&D at Zanskar
As Head of Agentic Systems R&D at Zanskar, I lead research on durable AI agent systems for long-running scientific, technical, and operational work. My earlier work was in computational geoscience, spanning geothermal exploration, hydrologic modeling, geophysical methods, scientific software, field evidence, and uncertainty.
Core Expertise
Agent systems and evaluation · Geothermal exploration · Hydrogeological and geophysical modeling · Bayesian and machine-learning methods for uncertainty · Scientific software and reproducible field data
Selected publications
A selection; the full list is on the publications page.
Open Source Software & Resources
Software and resources from my geoscience research.
- SKBEL — Bayesian Evidential Learning framework — Historical
- MGS-public — Minimum Gradient Support inversion for resistivity/IP data
- pysgems — Use SGeMS (Stanford Geostatistical Modeling Software) within Python — Deprecated
- Python Scientific Project Template — Standardized structure for scientific Python projects
Datasets
Public datasets and field-data packages.
- Lightning Dock Geothermal Field Data Package — 2026, Geothermal Data Repository
- Zanskar’s GeoGym — Benchmark dataset for evaluating geothermal exploration strategies
- 4D ERT Monitoring Dataset — Temperature monitoring with geophysical data
- Wellhead Protection Area Prediction Dataset — Groundwater modeling and protection
Writing
Essays, tutorials, and research notes.
- Comparing Well and Geophysical Data for Temperature Monitoring Within a Bayesian Experimental Design Framework — October 19, 2022
- Principal Component Analysis from the ground up with Python — May 03, 2022
- A Gentle Introduction to Python with PyCharm — May 01, 2022
GitHub Stats
news
| Oct 02, 2024 | SLC Bound! 🏔️ |
|---|---|
| Aug 02, 2023 | Berkeley Bound! 🌁 |
| Mar 28, 2023 | PhD defended! |
| Mar 09, 2023 | Public PhD defense. "Machine Learning for Bayesian Experimental Design in the Subsurface" |