Get Started
Hunting for you, 24x7Get the iPhone app

Your hunt runs in your pocket, and tells you the moment a strong match opens.

Download on theApp Store
WeekdayPosted in November 2026

ML Infrastructure and Performance Engineer, AI Expert

AnywhereMid levelRemoteFull-time$90 – $120 an hour
Who we'd reach at WeekdayFound and reached by Power Apply
Best leadHiring manager

Leads the team this role joins

Team lead

Runs the team day to day

Founder

Decides on early hires

Power Apply

From your own Gmail and LinkedIn.

About the role

This is a fully remote, full-time AI evaluation role open to engineers worldwide with at least 2 years of ML systems experience. It suits someone who can write hard performance and infrastructure tasks and grade AI-generated solutions. Weekday is a hiring platform that places engineers with client companies. Engagement is arranged through a local partner based on where you live.

What you’ll do

  • Design technical tasks on accelerator kernels, profiling, distributed-workload debugging and inference serving.
  • Write difficult, adversarial engineering problems to test advanced AI models.
  • Produce reference solutions that follow ML systems best practice.
  • Grade AI-generated solutions for correctness, efficiency, scalability and practicality.
  • Write guidelines and rubrics for kernel optimisation, profiler analysis, distributed systems and inference performance.
  • Find bottlenecks across ML infrastructure, compilers, accelerators and model serving.
  • Give structured feedback to improve model capability in these areas.
  • Work with researchers to keep training and evaluation data high quality.

What they’re looking for

  • At least 2 years of hands-on work in ML systems, ML infrastructure, model serving or accelerator performance.
  • Practical experience with kernel programming, profiling, distributed debugging or LLM inference serving.
  • Production experience with JAX and or PyTorch.
  • Experience with tools such as Pallas, Triton, CUDA, CUTLASS, JAX/XLA Profiler, torch.profiler, NVIDIA Nsight, vLLM, SGLang, TensorRT-LLM or JetStream.
  • Understanding of accelerators such as TPUs, NVIDIA H100 or B200, AWS Trainium and AMD MI300.
  • Ability to analyse and improve throughput, latency, memory use and efficiency.
  • Clear written communication of technical decisions.
  • Availability for an exclusive 40-hour weekday commitment with no conflicting engagements.

Nice to have

  • Experience with JAX, XLA, TPU environments, custom operators, sharding and distributed training.
  • Experience with compiler technology, performance benchmarking or AI model evaluation.

What’s on offer

  • USD 90 to 120 per hour.
  • Fully remote with worldwide eligibility.
  • Engagement or employment runs through a local partner based on country of residence.

Skills

JAXPyTorchTritonCUDACUTLASSvLLMSGLangTensorRT-LLMJetStreamNVIDIA NsightKernel OptimisationPerformance ProfilingDistributed TrainingLLM InferenceTPUModel Serving

Questions about this role

Is the role remote?

Yes. It is fully remote and open worldwide.

What is the pay?

USD 90 to 120 per hour.

How much experience is needed?

At least 2 years of hands-on professional experience in a relevant area.

Is it full time?

Yes. It asks for an exclusive 40-hour weekday commitment.