Member of Technical Staff, ML Systems

Job Description

Job Description

Member of Technical Staff, ML Systems

Company: Confidential — venture-backed technology company
Location: Menlo Park, CA (in office 5 days per week)
Compensation: $180,000 - $230,000 + competitive equity
Employment Type: Full-time
Visa Sponsorship: Visa transfers and new sponsorship (H-1B, TN)

About the Company

The client builds the training and inference stack for world models. Today's stack was built for language models; the client is rebuilding it for video, image and world-model workloads by co-designing low-level GPU kernels, distributed systems and the models themselves. Its public benchmarks include MiniMax running roughly ten times faster at half the cost, 2K image generation in about four seconds for three cents, and LTX video running faster than real time.

The client has five founders covering distributed systems, GPU kernel optimization, cloud infrastructure and research, with prior work at Fireworks AI, Meta, Google, Apple, Microsoft, Snowflake and Alibaba. It has raised a $10M seed.

The Role

The client is hiring ML systems engineers to own speed and efficiency across its stack: low-level kernels, distributed inference engines and multi-node training and serving systems. You report directly to the cofounder and CEO. The work sits below the application layer (kernels, runtimes and distributed engines for video and world models), so this is not an agents or RAG role. You will feel at home if you would rather make a video model ten times faster than train one.

What You Will Do

• Optimize GPU and system performance for training and inference on image, video and world-model workloads.
• Profile and remove bottlenecks at the kernel, memory, system and cluster level using Nsight and related tooling.
• Write low-level optimizations in CUDA and Triton on code paths that run in production.
• Build distributed inference and training engines for diffusion models across multiple GPUs and nodes.
• Own communication performance: NCCL, RDMA over InfiniBand or RoCE, and disaggregated serving.
• Build benchmarking and regression harnesses so performance gains hold in production.

What You Bring

• Authored core features in an inference or training framework (vLLM, SGLang, TensorRT-LLM, Megatron or similar), not only deployment or integration work
• 1+ year of full-time, hands-on ML systems or GPU performance work in your current or most recent role
• Kernel work on NVIDIA GPUs in CUDA, CUTLASS, Triton or PTX
• Inference or training performance experience: GPU kernels, runtimes, distributed execution
• Strong ML and CS fundamentals; degree in CS or a related quantitative field
• Ability to work in the Menlo Park office 5 days a week

Nice to Have

• Optimized diffusion, video, image or other multimodal workloads
• Public, verifiable kernel contributions to open-source ML systems

Interview Process

Hiring manager screen with the CEO (30 min), domain deep dive (60 min), system design (60 min), optional onsite.

Tech Stack

CUDA, Triton, PyTorch, Nsight, NCCL, RDMA

Before you apply: get a free recruiter-readiness score for your resume at CareerStanding — about a minute, no signup, and your resume stays in your browser.