AI Inference Engineer: Kernel & Edge Optimization

Lever, Inc.; Town of Italy (NY); On-site; EUR 90,000 - 130,000; Full time

Benefits offered by this job

Remote-first environment

International team

Exposure to cutting-edge AI research

Performance-critical infrastructure

Job summary

Lever, Inc. is seeking an AI Research Engineer (Kernel & Inference Optimization) based in Italy.

You will work at the intersection of AI research, systems engineering, and high-performance model inference, developing and optimizing model-serving architectures for diverse hardware environments. You will tackle latency, throughput, memory efficiency and scalability challenges across mobile and edge platforms, including GPU kernel work and advanced optimization techniques.

Qualifications

• PhD in NLP/ML or related field or strong AI research track record.
• Proven expertise in Metal Shading Language (MSL) and writing custom compute shaders.
• Experience with low-level kernel optimization and inference optimization on mobile or constrained devices.
• Track record of improving inference latency, throughput and memory footprint.
• Strong understanding of model-serving architectures and optimization techniques.
• Experience developing GPU kernels for mobile devices.
• Ability to run empirical research and systematic experiments to solve bottlenecks.

Responsibilities

• Design and deploy model-serving architectures optimized for throughput and latency.
• Develop inference pipelines for mobile and edge platforms.
• Set performance targets for latency, speed, throughput, memory and reliability.
• Create benchmarks and collect metrics on latency, throughput and errors.
• Build datasets and simulations for evaluating performance.
• Identify bottlenecks and implement batching, memory management and networking optimizations.
• Develop custom GPU kernels and compute shaders for mobile hardware (MSL).
• Apply pruning, quantization, Flash Attention, KV caching and speculative decoding.
• Design distributed inference with tensor/pipeline/expert parallelism for large GPUs.
• Collaborate to integrate optimized frameworks into production and edge devices.
• Define evaluation methodologies and report experiments against benchmarks.
• Monitor production performance for scalability and reliability improvements.

Skills

MSL programming

GPU kernel optimization

Inference optimization

Tensor/parallels

Diffusion models

Vision Transformers

Edge/mobile deployment

Benchmarking

Education

PhD in NLP/ML or related

Degree in Computer Science

Tools

TensorRT/ONNX Runtime

MLIR

CoreML/Metal tooling

Job description

Lever, Inc. is seeking an AI Research Engineer (Kernel & Inference Optimization) based in Italy.

You will work at the intersection of AI research, systems engineering, and high-performance model inference, developing and optimizing model-serving architectures for diverse hardware environments. You will tackle latency, throughput, memory efficiency and scalability challenges across mobile and edge platforms, including GPU kernel work and advanced optimization techniques.