ML Research Engineer - PyTorch Expert

Mercor; Toronto; On-site; CAD 120,000 - 170,000; Full time

Job summary

Mercor seeks an ML research engineer to generate high-impact ideas for long-horizon AI research projects. You will design hard tasks to challenge frontier models, focusing on inference, evaluation, and systems work more than bulk training.

You will explain why each problem is hard for frontier models and collaborate with a leading AI lab. PhD or MSc with strong research output preferred.

Qualifications

• Current or recent role as an ML research engineer, research scientist or applied scientist.
• Hands-on experience training, fine-tuning or optimizing models in PyTorch or JAX.
• Deep knowledge of LLMs, GPU inference and performance, RL, or Scientific ML.
• Ability to explain why a problem is hard for a frontier model.

Responsibilities

• Come up with ideas for hard machine learning engineering tasks for a long-horizon AI research project.
• Design tasks hard enough to stump frontier models: a strong task solved in no more than 2 of 8 attempts.
• Lean toward inference, evaluation and systems work rather than heavy model training.
• Explain clearly why each problem is hard for a frontier model.
• You mainly contribute ideas. You do not need to build or review the tasks yourself.

Skills

ML research role

PyTorch/JAX hands-on

LLMs

GPU inference

Reinforcement learning

Scientific ML

Explain frontier-model hardness

Education

PhD in ML/CS/Physics

MSc with strong research output

Tools

PyTorch

JAX

CUDA

Triton

vLLM

Job description

To qualify you must work as an ML research engineer, research scientist or applied scientist, and have personally trained, fine-tuned or optimized models in PyTorch or JAX.

What you'll do

• Come up with ideas for hard machine learning engineering tasks ("plans") for a long-horizon AI research project with a leading AI lab
• Design tasks hard enough to stump frontier models: a strong task is solved in no more than 2 out of 8 attempts
• Lean toward inference, evaluation and systems work rather than heavy model training. About half of the tasks run on GPU and half are CPU-only
• Explain clearly why each problem is hard for a frontier model
• You mainly contribute ideas. You do not need to build or review the tasks yourself

Must-haves

• Current or recent role as an ML research engineer, research scientist or applied scientist
• Hands-on experience personally training, fine-tuning or optimizing models in PyTorch or JAX
• Deep knowledge of at least one of these areas:
  • LLMs: pretraining, alignment, evals
  • GPU inference and performance: vLLM, Triton, CUDA, profiling
  • Reinforcement learning (RL)
  • Scientific ML: weather and climate, physics, biology
• Able to explain why a problem is hard for a frontier model

Nice-to-haves

• Experience writing benchmark or evaluation questions (a big plus)
• PhD in ML, CS or physics, or an MSc plus strong research output
• Publications
• Prior Mercor contracts with strong performance

Not a fit

• Software or full-stack engineers whose AI experience is mainly building LangChain or RAG applications
• Backgrounds limited to AI trainer, rater or annotator work