GPU Programming Software Engineer

Design and implement high-performance GPU software and task pipelines to support training large language models, with a strong focus on kernel and shader performance, architectural efficiency, and host-side integrations.

• Design, implement, and optimize GPU-based software solutions using CUDA, WebGPU, or GLSL.
• Profile and fine-tune GPU kernels and shaders to maximize performance and efficiency on target hardware.
• Develop host-side logic and integrations in C++ to support GPU workloads and data flows.
• Create GPU-related tasks and deliverables that meet the project specifications for LLM training workflows.
• Deliver output that passes minimum submission requirements for the project, with attention to performance and correctness.

• Required skills: CUDA, WebGPU, GLSL, C++.
• Preferred experience with GPU programming languages and toolchains that run on NVIDIA hardware.
• Background in one or more of the following, relevant to GPU work, is strongly preferred, graphics programming, ML acceleration, scientific computing, high performance computing, or other domains using GPU technologies.
• Proven ability to profile and optimize GPU kernels and host integrations for production workloads.

• Engagement type, contractor.
• Location, remote.
• Output-based work, experts are paid per task that meets project specifications.
• There is a minimum weekly task submission requirement; the source did not specify the exact number.
• Time required to complete each task will vary by experience and workflow.

• Stated pay range, $60 to $85 per hour.
• Primary compensation structure is output-based, paid per completed task that meets project specifications.

• Roles are typically filled within 48 hours, and candidates available to start immediately are preferred.
• If selected, you will be expected to begin your first tasks within 24 to 48 hours after completing onboarding.

micro1 is an AI data lab focused on training frontier models and evaluating AI agents, converting real-world subject matter expertise into high-quality training data and evaluation tasks across domains such as finance, healthcare, and engineering.