AI Inference Architect: Kernel & Edge Optimizations

Lever, Inc.; United Arab Emirates; Remote; AED 350,000 - 700,000; Full time

Benefits offered by this job

Remote-first

International team

Cutting-edge research

Performance infrastructure

Collaboration emphasis

Job summary

Lever, Inc. in United Arab Emirates is seeking an AI Research Engineer (Kernel & Inference Optimization) to work on model-serving architectures across hardware, from mobile to edge devices. You will benchmark, optimize, and deploy inference pipelines, with emphasis on latency, throughput, memory efficiency, and scalable GPU kernels.

Remote collaboration with a global team will be central to the role. PhD preferred; strong experience in low-level optimization and production-ready inference is a

Qualifications

• PhD in NLP/ML with strong AI research track record.
• Proven expertise in Metal Shading Language and custom compute shaders.
• Experience optimizing inference on mobile or constrained devices.
• Track record of measurable improvements in latency, throughput, and memory footprint.

Responsibilities

• Design model-serving architectures with high throughput and low latency.
• Develop inference pipelines for mobile and edge devices.
• Set performance targets for latency and throughput.
• Build controlled benchmarks and monitor memory usage.
• Create datasets and scenarios for evaluating performance.
• Identify bottlenecks and implement system-level optimizations.
• Develop GPU kernels and compute shaders for mobile hardware.
• Collaborate with cross-functional teams to productionize optimized frameworks.

Skills

MSL

GPU Kernels

Low-level Opt

Model Serving

Benchmarking

C/C++

Python

Education

PhD in NLP/ML

Tools

CUDA

TensorRT

Mobile GPUs

Job description

Lever, Inc. in United Arab Emirates is seeking an AI Research Engineer (Kernel & Inference Optimization) to work on model-serving architectures across hardware, from mobile to edge devices. You will benchmark, optimize, and deploy inference pipelines, with emphasis on latency, throughput, memory efficiency, and scalable GPU kernels.

Remote collaboration with a global team will be central to the role. PhD preferred; strong experience in low-level optimization and production-ready inference is a