Terra AI jobs

Terra AI is building an AI platform for critical resource exploration, combining machine learning with geoscience to help mining companies discover copper, lithium, and other strategic minerals faster and more efficiently. The company is seeking a deep learning engineer or scientist to lead development of a generative diffusion model that produces 3D geological models conditioned on geophysical surveys and borehole data, enabling explorers to make better-informed subsurface decisions. Key responsibilities include designing and training diffusion models for 3D geological generation, conditioning models on geophysical observations, generating synthetic training data, and adapting approaches for real-world projects in collaboration with teams. Required expertise includes extensive PyTorch experience, proven ability to develop large deep learning models from scratch, hands-on data curation skills, and strong software engineering fundamentals including version control and code optimization. Nice-to-haves include familiarity with diffusion models and posterior sampling, transformer architectures for 3D data, distributed GPU training, and cloud ML infrastructure. Terra AI is backed by Khosla Ventures and strategic partners like Rio Tinto, positioning the role as an opportunity to apply cutting-edge generative AI to reshape how the world discovers critical resources for the energy transition.

Terra AI, founded to revolutionize critical resource exploration through AI and geoscience, is seeking a deep learning engineer or scientist to lead development of a generative model that produces 3D geological models from geophysical surveys and borehole data—analogous to how image generators work but for subsurface modeling. You'll design, train, and iterate on diffusion models for geological applications, condition generation on real-world observational data, create synthetic training data, and collaborate with project teams to adapt the approach to specific exploration challenges. Required qualifications include extensive PyTorch experience, proven ability to develop large deep learning models from scratch, hands-on data curation skills, and strong software engineering practices; nice-to-haves are familiarity with generative/diffusion models, transformers (especially for 3D), distributed GPU training, and cloud ML infrastructure. The company is backed by Khosla Ventures and works with major industry partners like Rio Tinto, operating at the intersection of clean energy transition and critical mineral discovery. This is a technically deep role focused on building a foundational generative model rather than applied data science, requiring someone who can architect novel deep learning systems from the ground up.

Terra AI is building AI systems to accelerate critical resource exploration for mining and subsurface industries, combining machine learning with geoscience expertise. This role sits within the engineering team tasked with productionizing internal modeling tools—taking scientific workflows and transforming them into robust, maintainable, extensible software that serves geophysics and reservoir simulation domains. You'll collaborate with domain experts to refactor modeling stacks, design APIs, implement testing strategies suited to scientific computing, and build orchestration pipelines for simulation ensembles, working primarily in Python and Julia. Required skills include strong software engineering fundamentals, production-grade Python, willingness to work in Julia, experience with ML frameworks at integration level (PyTorch preferred), orchestration tools (Flyte, Prefect, Dagster), and familiarity with performance profiling and workflow design. Geophysics, reservoir simulation, PDE-based HPC work, or prior experience supporting scientific users are valuable but not required. Terra AI is backed by Khosla Ventures with industry partnerships including Rio Tinto and operates at the intersection of AI and critical minerals, making this a meaningful role in the energy transition space.