Staff ML Performance Engineer — Edge Inference Optimizer

Icehouseventures; Greater London; Hybrid; GBP 70,000 - 90,000; Full time

Job summary

Icehouseventures is looking for a Staff ML Performance Engineer to optimize ML inference for edge accelerators and GPUs. You will focus on running large transformer-based models efficiently on low-cost, low-power devices, directing technical projects and ensuring reliable operation in vehicular compute environments.

This full-time role operates under a hybrid policy, balancing office collaboration with remote work. Candidates can expect a dynamic environment, fostering innovation and inclusion.

Qualifications

• Proven experience improving performance in production systems with tight constraints.
• Strong proficiency with at least one relevant stack/toolchain.
• Comfort operating at multiple levels of abstraction.
• Strong software engineering fundamentals.
• Clear communicator and collaborative teammate.

Responsibilities

• Profile and pinpoint bottlenecks across the inference stack.
• Implement and validate optimisations in compilers and runtimes.
• Build robust benchmarking and regression testing.
• Optimise for multiple targets.
• Collaborate with model developers.

Skills

Performance improvement in production systems

Proficiency with TensorRT, CUDA, Qualcomm QNN

Software engineering fundamentals

Collaboration and communication skills

Python proficiency

C++ proficiency

Tools

TensorRT

CUDA

Qualcomm QNN

Triton

OpenCL

Job description

Icehouseventures is looking for a Staff ML Performance Engineer to optimize ML inference for edge accelerators and GPUs. You will focus on running large transformer-based models efficiently on low-cost, low-power devices, directing technical projects and ensuring reliable operation in vehicular compute environments.

This full-time role operates under a hybrid policy, balancing office collaboration with remote work. Candidates can expect a dynamic environment, fostering innovation and inclusion.