Back to jobs

Member of Technical Staff - ML Performance

Toronto · Zürich · Seattle · California · Toronto, Ontario, Canada

Actively hiring, first seen by us 14 hours ago
Pay
Pay not listed
Work setting
Hybrid — work setting source
Location type: Hybrid
From the employer’s posting
Employment
Full-time — employment source
Employment type: Full-time
From the employer’s posting
Team
Technical Staff — team source
Department: Technical Staff
From the employer’s posting

What you’ll bring

All qualifications

Core experience

  • Bachelor's degree or equivalent hands-on experience in Computer Science, Computer Engineering, or a related technical field.
  • Experience optimizing training or inference workloads across very large GPU clusters, including topology-aware placement, scaling efficiency, performance isolation, and diagnosing failures that emerge only at fleet scale.
  • Experience profiling live training and inference runs with Nsight Systems or the PyTorch profiler and translating traces into quantifiable step-time or MFU improvements.
  • Experience writing Triton, CUTLASS, or CuTe-DSL kernels, or contributing to open-source kernel libraries.
  • Expertise in at least one of low-precision numerics, kernel authoring, or large-run fault diagnosis, and credibility in the others.
  • Experience implementing context or sequence parallelism for long-horizon video or high-token-count models.
Qualification wording
Bachelor's degree or equivalent hands-on experience in Computer Science, Computer Engineering, or a related technical field.
Experience optimizing training or inference workloads across very large GPU clusters, including topology-aware placement, scaling efficiency, performance isolation, and diagnosing failures that emerge only at fleet scale.
Experience profiling live training and inference runs with Nsight Systems or the PyTorch profiler and translating traces into quantifiable step-time or MFU improvements.
Experience writing Triton, CUTLASS, or CuTe-DSL kernels, or contributing to open-source kernel libraries.
Expertise in at least one of low-precision numerics, kernel authoring, or large-run fault diagnosis, and credibility in the others.
Experience implementing context or sequence parallelism for long-horizon video or high-token-count models.
Education & alternatives
Requirements - Bachelor's degree or equivalent hands-on experience in Computer Science, Computer Engineering, or a related technical field. - Deep hands-on experience with PyTorch and at least one large-scale parallelism stack (FSDP2, Megatron-Core, TorchTitan, or DeepSpeed) on real multi-node jobs, not single-node approximations.

About Veeda AI

Veeda AI is building the next generation of multimodal foundation world models for Physical AI.

In the employer’s words · Read in context · Company website ↗

Apply at Veeda AI

Posting record

Report an error

Actively hiringFirst seen by us today, still listed, and not a repost of an earlier posting.

Status in our records
Live
First seen by us
Nov 18, 2026today
Recorded sightings
2
Last seen by us
Nov 19, 2026
Employer says posted
Aug 19, 2026

These dates show when we found the posting. Check the employer’s website to confirm it is still accepting applications.

How this job compares

Time since we first saw it

Day 0

Early. Similar jobs usually come down between day 7 and day 76.

From 5,990 similar postings that came down. Coming down is not always a hire.

Veeda AI is hiring

3 open jobs

Open jobs by the week we first saw them, last 12 weeks.

Hiring record

See how this role fits your experience

Add your resume and we’ll mark which of these tools you already use, and point out anything in the posting that could rule you out.

  • C++
  • Python
  • PyTorch
Where the posting names them
- Deep hands-on experience with PyTorch and at least one large-scale parallelism stack (FSDP2, Megatron-Core, TorchTitan, or DeepSpeed) on real multi-node jobs, not single-node approximations. - Fluency in Python and C++/CUDA with the ability to predict where a kernel will stall from its memory access pattern before profiling. - Experience profiling live training and inference runs with Nsight Systems or the PyTorch profiler and translating traces into quantifiable step-time or MFU improvements.
Responsibilities - Distributed Training Throughput: Own step time and model FLOPs utilization for multi-node video world model training, choosing the tensor, context, and expert parallelism mix in PyTorch FSDP2 and Megatron-Core rather than inheriting a default. - Inference Throughput Optimization: Profile end-to-end latency, GPU utilization, memory pressure, and kernel efficiency with PyTorch Profiler, Nsight Systems, Nsight Compute, and torch.utils.benchmark, then improve throughput through torch.compile/Inductor, CUDA Graphs, mixed and low precision, quantization, operator fusion, and multi-GPU serving.
- Distributed Training Throughput: Own step time and model FLOPs utilization for multi-node video world model training, choosing the tensor, context, and expert parallelism mix in PyTorch FSDP2 and Megatron-Core rather than inheriting a default. - Inference Throughput Optimization: Profile end-to-end latency, GPU utilization, memory pressure, and kernel efficiency with PyTorch Profiler, Nsight Systems, Nsight Compute, and torch.utils.benchmark, then improve throughput through torch.compile/Inductor, CUDA Graphs, mixed and low precision, quantization, operator fusion, and multi-GPU serving. - Precision & Numerical Stability: Take BF16, FP8, and NVFP4 recipes from running to converging on Blackwell, chasing scaling-factor and accumulation bugs into the video tokenizer and VAE layers where the activation outliers actually live.
- Precision & Numerical Stability: Take BF16, FP8, and NVFP4 recipes from running to converging on Blackwell, chasing scaling-factor and accumulation bugs into the video tokenizer and VAE layers where the activation outliers actually live. - Kernels & Compilation: Write and tune the CUDA and Triton kernels PyTorch does not give us, driving FlashAttention-4, FlexAttention, and torch.compile integration so quadratic attention over long video sequences stops setting step time. - Communication & Overlap: Tune NCCL collectives and compute/communication overlap across NVLink domains and the fabric, using the NCCL flight recorder to turn a watchdog timeout into a named rank and collective, not a restart.
- Bachelor's degree or equivalent hands-on experience in Computer Science, Computer Engineering, or a related technical field. - Deep hands-on experience with PyTorch and at least one large-scale parallelism stack (FSDP2, Megatron-Core, TorchTitan, or DeepSpeed) on real multi-node jobs, not single-node approximations. - Fluency in Python and C++/CUDA with the ability to predict where a kernel will stall from its memory access pattern before profiling.
- Fluency in Python and C++/CUDA with the ability to predict where a kernel will stall from its memory access pattern before profiling. - Experience profiling live training and inference runs with Nsight Systems or the PyTorch profiler and translating traces into quantifiable step-time or MFU improvements. - Expertise in at least one of low-precision numerics, kernel authoring, or large-run fault diagnosis, and credibility in the others.

Your next step

Complete your application on jobs.ashbyhq.com. The employer’s form will show what is required.

Applying through Ashby

  • No account needed. “Autofill from resume” at the top of the form fills the key fields from your file.

Already applied? Track this application

Job description

View original posting ↗

Member of Technical Staff - ML Performance

About Us

Veeda AI is building the next generation of multimodal foundation world models for Physical AI. We're a small, fast-moving team of engineers and researchers from leading AI labs, tackling some of the most challenging problems at the intersection of AI, robotics, and embodied intelligence. If you're excited about pushing the boundaries of what's possible with Physical AI, you'll have the opportunity to make an outsized impact from day one.

Responsibilities

  • Distributed Training Throughput: Own step time and model FLOPs utilization for multi-node video world model training, choosing the tensor, context, and expert parallelism mix in PyTorch FSDP2 and Megatron-Core rather than inheriting a default.

  • Inference Throughput Optimization: Profile end-to-end latency, GPU utilization, memory pressure, and kernel efficiency with PyTorch Profiler, Nsight Systems, Nsight Compute, and torch.utils.benchmark, then improve throughput through torch.compile/Inductor, CUDA Graphs, mixed and low precision, quantization, operator fusion, and multi-GPU serving.

  • Precision & Numerical Stability: Take BF16, FP8, and NVFP4 recipes from running to converging on Blackwell, chasing scaling-factor and accumulation bugs into the video tokenizer and VAE layers where the activation outliers actually live.

  • Kernels & Compilation: Write and tune the CUDA and Triton kernels PyTorch does not give us, driving FlashAttention-4, FlexAttention, and torch.compile integration so quadratic attention over long video sequences stops setting step time.

  • Communication & Overlap: Tune NCCL collectives and compute/communication overlap across NVLink domains and the fabric, using the NCCL flight recorder to turn a watchdog timeout into a named rank and collective, not a restart.

  • Fault Diagnostics & Recovery: Build the detection layer for silent data corruption (SDC), stuck CUDA kernels, and "card-freeze" hangs, plus asynchronous and tiered checkpointing that makes an interruption cost minutes rather than a day.

Requirements

  • Bachelor's degree or equivalent hands-on experience in Computer Science, Computer Engineering, or a related technical field.

  • Deep hands-on experience with PyTorch and at least one large-scale parallelism stack (FSDP2, Megatron-Core, TorchTitan, or DeepSpeed) on real multi-node jobs, not single-node approximations.

  • Fluency in Python and C++/CUDA with the ability to predict where a kernel will stall from its memory access pattern before profiling.

  • Experience profiling live training and inference runs with Nsight Systems or the PyTorch profiler and translating traces into quantifiable step-time or MFU improvements.

  • Expertise in at least one of low-precision numerics, kernel authoring, or large-run fault diagnosis, and credibility in the others.

Nice to Have

  • Experience optimizing training or inference workloads across very large GPU clusters, including topology-aware placement, scaling efficiency, performance isolation, and diagnosing failures that emerge only at fleet scale.

  • Experience writing Triton, CUTLASS, or CuTe-DSL kernels, or contributing to open-source kernel libraries.

  • Experience implementing context or sequence parallelism for long-horizon video or high-token-count models.

  • Experience running or porting large training workloads on AMD GPUs (ROCm) or Google TPUs (JAX/XLA).

  • Experience building fault-tolerant training with elastic world size, dynamic node re-queueing, or asynchronous distributed checkpointing.

  • Experience optimizing generative inference for interactive rollouts, including few-step samplers, distillation, and KV or latent caching.

  • Publications or presentations on machine learning systems, compilers, or high-performance kernels.

Demand for ML Engineer roles

New ML Engineer postings at the employers we follow are up 12% over the last 6 weeks compared with the 6 before.

Weekly new ML Engineer postings at the same employers, from the week of August 17 to the week of November 9. The fewest in a week was 556, the most 1,080, and the latest week had 556.
August 17Last week

Each bar is one week of new postings. The darkest bars are the last 6 weeks.

Counted from the hiring pages we’ve followed for at least five weeks, so employers we add don’t look like growth. Updated weekly.

About the role

We file this job under ML Engineer, by the work the posting describes.

ML engineers put machine learning models into products. They write the code that trains, ships, serves and monitors models so a recommendation or prediction works for real users.

Postings ask for strong Python and software engineering, a framework such as PyTorch, cloud services, and more and more often experience with large language models.

A data scientist often works out which model to build; an ML engineer makes it run reliably in production.

US government figures for computer and information research scientists

Median yearly pay, May 2025
$140,300
Expected change, 2025 to 2035
21.8% more jobsagainst 3.5% for all jobs

The government has no code for ML engineers, so these figures are for the closest group, computer and information research scientists.

Source: US Bureau of Labor Statistics, Occupational Employment and Wage Statistics (May 2025) and Employment Projections (2025 to 2035). Occupation code 15-1221.

Source details

View original posting ↗

Source notes

Source excerpts

Selected passages from the saved posting. Check the full description for conditions and exceptions.

Pay

No pay amount identified in the saved description.

Location & working pattern

Toronto, Ontario, Canada

Working pattern and location restrictions need checking in the full posting.

Work authorization

No clear work-authorization passage found. Eligibility is unconfirmed.

Find answers in the posting

AI
How answers work

AI selects complete passages from this posting. Check them for conditions and exceptions.

Uses this posting and your question. No profile needed.