Software Engineer, AI Inference

Skild AI

1d • $100,000 - $300,000

About The Position

We are looking for a Software Engineer to work at the forefront of deploying our cutting-edge AI models, enhancing the performance and capabilities of our embodied systems. You will be responsible for optimizing AI inference processes from lightweight to billion-parameter models, ensuring our robots operate with unmatched efficiency and intelligence in real-world environments. You will work at the intersection of systems and machine learning, directly contributing to making our AI models more powerful and adaptive by ensuring consistent performance in light of variable and perhaps unforeseen compute and hardware constraints.

Requirements

• BS, MS or higher degree in Computer Science, Robotics, Engineering or a related field, or equivalent practical experience.
• Minimum of 3 years of industry experience.
• Proficiency developing in low-level systems languages (C, C++, Rust, Go), Python and at least one deep learning library such as PyTorch, TensorFlow, JAX, etc.
• Deep understanding and practical experience with low-level systems concepts (multithreading, networking, embedded systems, memory management).

Nice To Haves

• Experience with CUDA.
• Deep understanding of state-of-the-art machine learning techniques and models.
• Experience optimizing various machine learning architectures.
• Experience with machine learning compilers.
• Experience optimizing model inference for robotic systems deployment.

Responsibilities

• Develop and optimize runtime AI inference pipelines for real-world robotic deployment.
• Build infrastructure, frameworks, and tooling to enable reliable integration of models into robotic systems and informative analysis of production models to drive the direction of architecture choice and deployment system design.
• Formulate specialized optimization solutions for various inference paradigms and scenarios (autoregressive models, denoising models, hierarchical models, state machines, multi-agent systems, cloud-based inference).
• Adapt optimization solutions to various compute, hardware, and networking constraints.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level