Machine Learning Infrastructure Engineer, Kernel Optimization

Efference

San Francisco (CA)

On-site

USD 180,000 - 240,000

Full time

14 days+

Get a reply from this employer

Your resume and cover letter, tailored to exactly what they’re hiring for.

Get past ATS filters

Job summary

Efference is seeking an engineer to optimize robot model inference on datacenter GPUs, reducing latency and compute cost. You will work from model-level profiling down to GPU kernels, ensuring behavior remains correct while accelerating performance.

You will profile workloads, implement CUDA/Triton kernels, and optimize inference graphs, with a focus on precision checks and regression testing across hardware and software versions.

Qualifications

  • Proven track record of improving GPU/ML workloads.
  • Strong CUDA, memory hierarchy, synchronization and parallel execution understanding.
  • Experience with PyTorch and C++/Python development.

Responsibilities

  • Profile perception and robot-policy workloads to identify kernel, memory, and CPU bottlenecks.
  • Implement and optimize CUDA or Triton kernels for expensive operators with fusion and tiling.
  • Improve inference graphs via compilation and memory traffic reduction.
  • Evaluate precision/quantization with researchers, monitoring accuracy and performance.
  • Benchmark with realistic batch sizes and latency targets, including tail latency.
  • Maintain correctness and performance regression tests across hardware/software versions.

Skills

CUDA programming
C++
Python
Profiling
ML frameworks (PyTorch)
Benchmarking & correctness

Tools

PyTorch
Triton
TensorRT
Profiling tools (Nsight, nvprof)

Job description

About the company

Efference provides high-performance perception and compute infrastructure for training and deploying robot models in the cloud. Our vision is to make robotic deployments scalable with shared compute.

We approach the problem by owning the entire path from the image sensor to the datacenter GPU. We shrink the load at the sensor, fix the uplink so robot traffic is fast and reliable, and clock synchronize the system so observations arrive time-aligned, ready for inference.

About the role

You'll make robot models run efficiently on datacenter GPUs, reducing inference latency and the compute cost of serving a fleet. You'll work from model-level profiling down to GPU kernels, verifying that each optimization preserves the behavior the robot depends on.

What you’ll do
  • Profile representative perception and robot-policy workloads, separating kernel execution, memory movement, synchronization, launch overhead, and CPU bottlenecks.
  • Implement and optimize CUDA or Triton kernels for expensive operators, using fusion, tiling, suitable layouts, and reduced memory traffic where they help.
  • Improve inference graphs and execution paths through compilation, graph capture, allocation reuse, and elimination of unnecessary host-device transfers.
  • Evaluate precision and quantization choices with researchers, checking numerical error and closed-loop task performance alongside speed and memory use.
  • Benchmark realistic batch sizes, sequence lengths, image shapes, and concurrent sessions with the orchestration team, including tail latency under contention.
  • Maintain correctness and performance regression tests across hardware and software versions, and upstream improvements where a shared library is the right home.
What we are looking for
  • Experience delivering measured performance improvements in GPU workloads or high-performance ML systems.
  • Strong CUDA or comparable GPU programming skills and understanding of memory hierarchies, occupancy, synchronization, and parallel execution.
  • Proficiency in C++ and Python and experience with PyTorch or another major ML framework.
  • Ability to use profiling tools to identify the limiting resource and choose an optimization that addresses it.
  • Careful benchmarking and correctness practice, including handling precision changes and separating microbenchmark gains from application gains.
Nice to Have
  • Experience with Triton, CUTLASS, TensorRT, compiler internals, or custom framework operators.
  • Work on attention, vision encoders, diffusion or flow models, and small-batch latency-sensitive inference.
  • Experience with multi-GPU communication, training kernels, model parallelism, or performance tuning for shared accelerators.
Get your free, confidential resume review.

or drag and drop your file here.

Similar jobs

Similar jobs worth comparing

Member of Technical Staff, ML Systems ML Systems
Member of Technical Staff, ML Systems ML Systems

TensorScale • San Francisco (CA)

On-site
USD 180,000 - 260,000
Machine Learning Infrastructure Engineer, Orchestration
Machine Learning Infrastructure Engineer, Orchestration

Efference • San Francisco (CA)

On-site
USD 180,000 - 240,000
AI Inference Kernel Optimization Engineer
AI Inference Kernel Optimization Engineer

Vidman AI • United States

Remote
USD 140,000 - 210,000
Machine Learning Engineer- Inference Optimization | Experienced Hire
Machine Learning Engineer- Inference Optimization | Experienced Hire

Susquehanna International Group, LLP • Bala Cynwyd (PA)

On-site
USD 150,000 - 230,000
ML Systems Engineer — Performance & Scale
ML Systems Engineer — Performance & Scale

Veeda Innovation • California

Hybrid
USD 150,000 - 230,000
Member of Technical Staff - ML Performance
Member of Technical Staff - ML Performance

Veeda Innovation • California

On-site
USD 150,000 - 230,000
Member of Technical Staff - ML Performance
Member of Technical Staff - ML Performance

Veeda AI • California

On-site
USD 180,000 - 240,000
Member of Technical Staff - ML Performance
Member of Technical Staff - ML Performance

Veeda AI • Seattle (WA)

On-site
USD 180,000 - 260,000
Training / AI Infrastructure
Training / AI Infrastructure

Genesis AI • United States

Hybrid
USD 190,000 - 240,000
Training / AI Infrastructure
Training / AI Infrastructure

Genesis • San Carlos (CA)

On-site
USD 180,000 - 240,000