Principal On-Device AI Engineer – Real-Time Inference

Unity Technologies; Mountain View (CA); On-site; USD 278,100 - 417,100; Full time

Benefits offered by this job

Employee stock ownership

Comprehensive health insurance

Generous vacation and personal days

Job summary

Unity Technologies is seeking a Principal Engineer for On-Device AI Inference & Systems to own the end-to-end inference and integration stack. You will optimize models for latency, memory, and power across mobile, desktop, and embedded hardware, and mentor a team to ship production-grade AI features inside Unity's engine.

You will collaborate with research and platform teams to deploy transformers and diffusion models using WebGPU, CoreML, and ONNX runtimes, while shaping engineering

Qualifications

• 8+ years in software/ML engineering with on-device / edge inference focus.
• Proven production deployment of transformer- and/or diffusion-based models on mobile/desktop hardware.
• Hands-on experience deploying models through WebGPU and related runtimes.

Responsibilities

• Own end-to-end optimization: export, graph transform, fusion, memory layout, kernel tuning.
• Drive integration between ML runtime and game engine for real-time performance.
• Lead and mentor a team of engineers; establish KPIs for latency, memory, and power.

Skills

On-device inference

Performance optimization

WebGPU

C++/GPU programming

Technical leadership

Education

Bachelor's degree in CS/ML

Tools

ONNX Runtime Web

Transformers.js

WebLLM

TensorFlow.js

Job description

Unity Technologies is seeking a Principal Engineer for On-Device AI Inference & Systems to own the end-to-end inference and integration stack. You will optimize models for latency, memory, and power across mobile, desktop, and embedded hardware, and mentor a team to ship production-grade AI features inside Unity's engine.

You will collaborate with research and platform teams to deploy transformers and diffusion models using WebGPU, CoreML, and ONNX runtimes, while shaping engineering