Remote GPU Kernel & Inference Engineer

Anyone Ai Inc.; Ciudad de Mendoza; Remote; ARS 199,855,000 - 307,470,000; Full time

Job summary

Tether is seeking an AI Research Engineer (Kernel & Inference Optimization) to join its globally distributed team, focusing on advanced model serving and low-latency inference strategies. This remote role spans devices from edge to cloud, requiring deep expertise in efficient serving pipelines and memory-aware optimizations for multi-modal AI workloads.

You will design, implement, and evaluate cutting-edge inference architectures, collaborate with cross-functional teams, and advance

Qualifications

• Deep expertise in model serving pipelines and inference frameworks.
• Experience with low-level kernel and inference optimizations on mobile/edge devices.
• PhD or publications in AI R&D are preferred.
• Proficiency with Metal Shading Language (MSL) and custom compute shaders.
• Strong understanding of Diffusion Models and Vision Transformers.
• Experience with parallelism strategies (Tensor/Pipeline/Expert).
• Ability to design robust evaluation and optimization frameworks.

Responsibilities

• Design and deploy state-of-the-art model serving architectures with high throughput and low latency.
• Build, run, and monitor inference tests in simulated and live production environments.
• Prepare high-quality test datasets and scenarios for low-resource devices.
• Analyze bottlenecks and memory usage to optimize serving pipelines.
• Collaborate with cross-functional teams to deploy optimized serving frameworks.

Skills

Model serving expertise

Low-latency optimization

GPU programming

Research mindset

Cross-functional collaboration

Education

PhD in NLP / Machine Learning or related field

Tools

Metal Shading Language (MSL)

GPU kernels

Tensor Parallelism

Pipeline Parallelism

Diffusion Models

Vision Transformers

Job description

Tether is seeking an AI Research Engineer (Kernel & Inference Optimization) to join its globally distributed team, focusing on advanced model serving and low-latency inference strategies. This remote role spans devices from edge to cloud, requiring deep expertise in efficient serving pipelines and memory-aware optimizations for multi-modal AI workloads.

You will design, implement, and evaluate cutting-edge inference architectures, collaborate with cross-functional teams, and advance