Lead Data Scientist

Role: Lead Data Scientist (Scientific Software Engineer / Computational Scientist) Location: Mountain View, CA (Hybrid – 3 days a week onsite) Contract Length: 9 months with possible extension H1b transfer also works Interview process: Screening Interview ➡️ Technical Interview ➡️ Manager Interview ➡️ Client Interview Experience Level: Lead Main Skills: • Python (NumPy/SciPy/CuPy) • C++ • PyTorch • Geostatistics • 3D Mathematics • CUDA/OpenMP • AI-assisted coding Short Overview: • Scientific Software Engineer or Computational Scientist with a niche background in scientific simulation, procedural generation, or computational physics. • This is an implementation-heavy role requiring a developer who can translate complex mathematical logic and generative ML models into performant code to solve high-dimensional geometric problems. Job Description: Simulation & Generative Modeling We're seeking a Simulation Engineer with deep expertise in scientific computing, procedural generation, or computational physics to build the core algorithms for our 3D subsurface modeling engine. The Role: This is an implementation-heavy position bridging procedural physics and generative ML. You'll translate complex mathematical logic and latent-space models into performant code, solving high-dimensional geometric problems at scale. What We're Looking For: Core Competencies: • Procedural Generation: Terrain synthesis, voxel engines, noise-driven systems • Scientific Computing: CFD, FEA, multi-physics solvers • Computational Geometry: 3D mesh processing, volumetric data structures, spatial partitioning Key Responsibilities: 1. Algorithmic Implementation — Design memory-efficient algorithms for massive 3D voxel arrays and sparse data structures; implement deterministic and stochastic geometric rules • Example: Build C++/Python kernels using 3D Perlin/Simplex noise and vector fields to simulate braided river systems • Example: Implement Boolean CSG algorithms for volumetric injections of igneous bodies 1. Generative ML Engineering — Architect and train models (GANs, Diffusion) for high-resolution 3D spatial data using PyTorch • Example: Generate realistic fracture networks via 3D generative models • Example: Apply neural style transfer to map sedimentary textures onto volumetric frameworks Required Technical Skills: • Languages: Expert Python (NumPy/SciPy/CuPy); proficient C++ for performance kernels • Mathematics: Linear algebra, vector calculus, coordinate transformations • ML Frameworks: PyTorch (generative AI, computer vision) • Performance: CUDA/OpenMP; parallel computing experience • Workflow: AI-assisted coding for rapid prototyping and testing Domain Knowledge: Mathematical maturity in: • Structural modeling (Boolean operations, volumetric intersections) • Sedimentology (layer stacking, erosion, flow simulation) • Tectonics (displacement fields, kinematic transformations) • Geostatistics (particle systems, stochastic models) Ideal Background: • MS/PhD in Computer Science, Applied Mathematics, Computational Physics, or equivalent • Portfolio/GitHub demonstrating procedural world-building, physics engines, or scientific simulators