Hiring Machine Learning Engineers, Research Scientists, and Performance/Compiler Engineers based in Canada to take part in a paid, remote 60-minute interview about your daily workflows, tools, and challenges in training, profiling, and deploying large-scale AI/ML models.
WHAT YOU'LL DO:
Take part in a 60-minute one-on-one interview (IDI) via video call, in English, with a UX/market researcher.
Share your hands-on experience with hardware (GPUs/TPUs), frameworks (PyTorch/JAX), and profiling/debugging tools.
Give feedback on concepts and prototypes in development (no live coding required).
ABOUT THE PROJECT:
This is the first in a series of studies. Selected participants may be invited to additional interviews between October and December 2026, subject to availability.
REQUIRED QUALIFICATIONS:
Currently residing and working in Canada.
Works at a company with 50+ employees (flexible for companies with 2–49 employees).
Role: ML Engineer, ML Developer, Research Scientist, ML Performance Engineer, or Compiler Engineer.
Minimum 2 years of recent hands-on experience developing, training, profiling, or deploying large-scale AI/ML models (PhD candidates in CS/MLSys also qualify).
Currently using NVIDIA GPUs (A100, H100, GB200) and/or Google TPUs (v4, v5e, v5p, v6e/Trillium).
Active user of PyTorch or JAX (or vLLM/XLA).
Regular use of at least one performance or debugging tool (e.g., NVIDIA Nsight Systems/Compute, PyTorch Profiler, Perfetto, TensorBoard, XProf, CUDA-GDB, or XMon).
Uses Google Cloud Platform, Amazon Web Services, or Microsoft Azure as your primary cloud platform.
Available for a 60-minute session in October 2026, during business hours (Canadian time zone).
WHAT WE OFFER:
Payment per completed session (amount shared after screening).
Fully remote participation, no travel required.
Possibility of continued engagement in future studies throughout Q4 2026.
Interested candidates should answer the screening questions below. Once selected, we'll send a scheduling link and a simplified NDA (we keep the end client's identity confidential).
