Senior AI Model Optimization Architect for Inference

Qualcomm; Ireland; On-site; EUR 120,000 - 180,000; Full time

Benefits offered by this job

Salary, stock and performance related—

Relocation and immigration support

Education Assistance

Life, Medical, Income and Travel保险

Job summary

QT Technologies Ireland Limited is seeking a Staff Engineer – AI Model Optimization Architect to lead end-to-end transformation and optimization for LLMs, VLMs, diffusion and multimodal models on Qualcomm accelerators. You will collaborate with compiler, performance, and accuracy teams to translate models into accelerator-efficient execution, balancing throughput, latency, memory and quality.

The role covers Day0 enablement through production deployment, with emphasis on scaling optimizations

Qualifications

• Expert level expertise in PyTorch and inference-focused model optimization.
• Hands-on experience with graph capture and compilation workflows (TorchDynamo, torch.compile).
• Deep understanding of transformer architectures, attention, MVEs, and performance trade-offs.

Responsibilities

• Architect and deliver model optimization strategies transforming PyTorch models for Qualcomm accelerators.
• Drive graph capture and deployment using PyTorch, ONNX, and torch.compile with model rewrites.
• Design and implement fusion kernels using DSL-based approaches (e.g., Triton).
• Co-design lowering strategies, kernel fusion, layout decisions, and runtime integration with compiler teams.
• Profile and optimize LLM/VLM/diffusion inference for throughput and latency across modes.

Skills

PyTorch

Python

TorchDynamo

ONNX

Triton

Model optimization

Inference engineering

Distributed systems

Education

Bachelor's degree in Engineering/CS/related (Bachelors)

Master's degree in Engineering/CS/related

PhD in Engineering/CS/related

Tools

TorchDynamo

ONNX

Triton

Job description

QT Technologies Ireland Limited is seeking a Staff Engineer – AI Model Optimization Architect to lead end-to-end transformation and optimization for LLMs, VLMs, diffusion and multimodal models on Qualcomm accelerators. You will collaborate with compiler, performance, and accuracy teams to translate models into accelerator-efficient execution, balancing throughput, latency, memory and quality.

The role covers Day0 enablement through production deployment, with emphasis on scaling optimizations