On-Device ML Engineer - Fast, Lean AI for Games

Unity; California; Hybrid; USD 218,000 - 284,000; Full time

Benefits offered by this job

Health insurance

Life insurance

Stock ownership

Retirement plans

Vacation days

Parental leave

Commute subsidy

Wellbeing programs

ERGs

Job summary

Unity in the United States is seeking a Senior Machine Learning Engineer for On-Device & Mobile AI to optimize and deploy multi-modal models so they run fast and within tight memory and power budgets on mobile and desktop devices.

You will own the end-to-end inference stack from training checkpoints through export, quantization, and kernel tuning, collaborating with WebGPU runtimes and native backends to ship features at interactive frame rates.

Responsibilities

• Inference & On-Device Optimization
• Own the optimization pipeline for the models you ship: model export, graph transformation, operator fusion, memory-layout planning, and hardware-specific tuning across NPU, mobile GPU, and desktop/laptop GPU.
• Apply quantization (INT4/INT8/FP16), weight sharing, pruning, and distillation to hit latency, memory, and power budgets.
• Write and tune WebGPU compute shaders and native kernels; profile with browser and platform tools; eliminate bottlenecks at op and memory bandwidth levels.
• Runtime & Systems Integration: work with WebGPU-targeted runtimes and native options; build glue code where needed; support on-device benchmarking in CI.

Skills

On-device inference

Transformer models

Diffusion models

WebGPU/WebGL

Python

C++/Swift/Obj-C

Profiling/Perf engineering

Collaboration

Tools

ONNX Runtime Web

CoreML

TFLite

ExecuTorch

WebGPU/WGSL

Metal

Vulkan

CUDA

Job description

Unity in the United States is seeking a Senior Machine Learning Engineer for On-Device & Mobile AI to optimize and deploy multi-modal models so they run fast and within tight memory and power budgets on mobile and desktop devices.

You will own the end-to-end inference stack from training checkpoints through export, quantization, and kernel tuning, collaborating with WebGPU runtimes and native backends to ship features at interactive frame rates.