ML Engineer (Internship and Full-time)

About the role

Tilde Research operates as an AI laboratory focused on mechanistic interpretability, model architecture design, and pretraining research. We are seeking engineers to build the systems and infrastructure required to push the limits of current machine learning capabilities.

Key facts

What you'll do

• Improve training and inference speeds for experimental model architectures.
• Develop and manage distributed systems designed for high-performance model training.
• Partner with research staff to turn theoretical concepts into practical gains in model control and interpretability.

Requirements

• Proven background in deep learning or relevant research fields.
• Demonstrated expertise in machine learning infrastructure, evidenced by open source projects, technical writing, or professional experience with large-scale training pipelines.
• Ability to architect and manage complete machine learning workflows.
• Strong verbal and written communication skills.
• Capacity for rapid technical learning.

Skills & tools

• PyTorch
• Jax
• Triton
• Tilelang
• TK

Practical notes

Please ensure your application highlights your specific contributions to infrastructure projects or research logs.