Geo-AI for geothermal exploration
Precision exploration
Workforce development
Geo-AI for geothermal exploration
Precision exploration
Workforce development
AI-Driven | Expert-Guided: The Next Generation of Geothermal
Turning Decades of Dormant Geological Data into Actionable Intelligence
Thermofilic LLC utilizes proprietary AI technology developed by co-founder Sebnem Duzgun, Ph.D., and funded by the U.S. Department of Energy to produce favorability mapping for conventional geothermal and subsurface AI modeling for EGS, that reduces exploration risk by half.
By collecting and curating hyper- and multi-spectral aerial and satellite images to extract geothermal indicators, such as minerals and surface temperature, our clients receive detailed favorability maps with confidence intervals that accelerate exploration timelines and nearly double drilling success rates.
Geoscience Education Reimagined for the Age of AI
To meet the growing global demand for reliable and cost-efficient resource discovery, we are partnering with universities in the United States—and soon internationally—to train geoscience students in data science and machine-learning methodologies that can feed directly into our AI platform.
This initiative not only scales our ability to deliver high-resolution, AI-assisted favorability mapping to industry and governments, but also helps cultivate a new generation of geoscientists equipped with the data science fluency essential to the future of geothermal exploration and operations.
From Research to Real-World Impact: Dr. Sebnem Duzgun’s Breakthrough in AI-Powered Exploration
Dr. Düzgün’s body of work provides the scientific and technical foundation for Thermofilic’s AI-driven exploration platform. The validated methodologies, large-scale datasets and demonstrated accuracy underpin the company’s mission of delivering risk-reduced, AI-powered geothermal discovery.
In 2019, Dr. Düzgün’s team at Colorado School of Mines was awarded funding by the U.S. Department of Energy (DOE) to apply machine-learning techniques to geothermal exploration. The project focused on using hyperspectral remote sensing data and combining it with geological and geophysical information to identify geothermal potential via surface manifestations.
The AI model was tested on the Brady, Desert Peak and COSO geothermal fields— achieving 92–95 % accuracy on independent datasets and 72–76 % when transferring the model to a differing site. This work marks one of the first fully documented applications of deep-learning in early-stage geothermal exploration, explicitly linking surface remote-sensing to subsurface potential.
A Platform for Partnership Across the Energy Value Chain
Thermofilic’s AI-powered precision-exploration technology opens the door to a vast range of partnership and customer opportunities across the global energy landscape. From geothermal developers and energy investors to research institutions and government agencies, our platform empowers every stage of resource discovery—reducing risk, accelerating timelines, and enabling data-driven decision-making. We’re actively seeking collaborations that extend our impact:
- Investors driving the energy transition and looking to scale proven, DOE-backed innovation.
- Industry partners—from utilities to exploration firms—seeking to enhance their geoscientific workflows through AI integration.
- Academic and research collaborators training the next generation of AI-enabled geoscientists.
- Technology and data providers who share our vision for smarter, connected exploration.
Together, we can transform how the world discovers and develops sustainable energy resources—bringing human expertise and artificial intelligence into perfect alignment.