Founding Engineer (Systems + ML)

European Recruitment BV; United States; On-site; USD 100,000 - 150,000; Full time

Job summary

European Recruitment BV is seeking a Founding Engineer to join their team in the United States. This critical role requires a generalist comfortable with all aspects of product development, from designing ML models to optimizing GPU performance and engaging with customers.

The ideal candidate will have up to 5 years of experience and be eager to take ownership of significant projects, improving existing systems and creating new features.

Qualifications

• 0-5 years of experience preferred.
• Experience in machine learning and GPU architecture is a plus.

Responsibilities

• Design and train new custom ML model architectures.
• Architect and build end-to-end high-performance pipelines.
• Engage with customers to understand their needs.
• Optimize GPU code and system components.
• Take ownership of features and systems.

Skills

Machine Learning

GPU Optimization

Rust

CUDA/Triton

Job description

We are looking for a Founding Engineer with 0-5 YOE to join our team as an extension of the founding team. We need a generalist who is hungry, hardworking, and capable of touching all parts of our product — from training custom ML models and optimizing GPU kernels to shipping compiler improvements and sitting in customer meetings. This hire is critical for us to scale and meet growing inbound demand.

Responsibilities

• Explore, design, and train new custom ML model architectures (e.g., diffusion models, physics-informed models) from scratch to achieve 1000x+ performance improvements for chip design.
• Architect and build end-to-end pipelines, including high-performance GPU kernels (CUDA/Triton), efficient data I/O in Rust, and latency-sensitive inference.
• Engage directly with customers in virtual and in-person meetings to understand their needs, deploy solutions, and gather feedback to drive product priorities.
• Continuously performance-optimize GPU code and other systems components, tackling bottlenecks as they arise to dramatically improve runtimes.
• Take ownership of large, ill-defined problems, rapidly iterating from zero to one on new features and systems, with the freedom to rebuild components as needed.