Senior Neural Rendering Software Engineer

Applied Intuition Applied Intuition · Robotics · Sunnyvale, CA · Autonomy Tooling Software Engineering

Senior Neural Rendering Software Engineer to own end-to-end development and optimization of neural rendering systems, including 3D Gaussian-based representations and other learned scene representations, across multiple products and sensor modalities. Design and scale GPU pipelines using CUDA for neural scene training, inference, and real-time rasterization. Architect and optimize multi-threaded C++ systems for scene management, streaming, scheduling, and CPU-side orchestration of neural and hybrid rendering workloads. Research, prototype, and productionize state-of-the-art neural rendering and reconstruction techniques.

What you'd actually do

  1. Own end-to-end development and optimization of neural rendering systems, including 3D Gaussian-based representations and other learned scene representations, across multiple products and sensor modalities
  2. Design and scale GPU pipelines using CUDA for neural scene training, inference, and real-time rasterization — including data layout, memory movement, synchronization, and kernel performance
  3. Architect and optimize multi-threaded C++ systems for scene management, streaming, scheduling, and CPU-side orchestration of neural and hybrid rendering workloads
  4. Research, prototype, and productionize state-of-the-art neural rendering and reconstruction techniques, keeping the engine at the frontier of the field
  5. Integrate neural rendering outputs with existing simulation and rendering infrastructure, including lightweight physically-based rendering (PBR) components where classical techniques complement neural approaches

Skills

Required

  • modern real-time rendering systems
  • neural graphics (e.g., 3D Gaussian-based representations)
  • high-performance multi-threaded C++/CUDA pipelines
  • neural scene simulation such as 3D Gaussian splatting, NeRFs, neural radiance fields, or similar learned volumetric/surface representations
  • CUDA and/or shader development
  • optimizing multi-threaded C++ applications
  • performance debugging, profiling, and system-level optimization across CPU and GPU

Nice to have

  • Publications or significant project contributions in neural rendering, novel view synthesis, or 3D reconstruction
  • Experience productionizing research-grade neural graphics systems into real-time engines
  • Familiarity with differentiable rendering or differentiable simulation frameworks
  • Experience designing or optimizing real-time ray tracing systems
  • Familiarity with autonomy-related sensors (cameras, LiDAR, radar, ultrasonic) and their simulation or rendering requirements

What the JD emphasized

  • neural rendering
  • 3D Gaussian-based representations
  • CUDA
  • real-time simulation
  • multi-threaded C++
  • GPU pipelines

Other signals

  • neural rendering
  • 3D Gaussian-based representations
  • CUDA
  • real-time simulation