Distributed AI Training Systems Engineer

Bison; San Francisco (CA); On-site; USD 400,000 - 500,000; Full time

Benefits offered by this job

Health benefits

Unlimited PTO

Parental leave

Relocation support

Job summary

Thinking Machines is hiring engineers to build scalable training systems that power frontier AI models. You’ll design, implement and operate distributed infrastructure across thousands of GPUs and nodes, with a focus on reliability and performance.

The role emphasizes deep understanding of training stacks, parallelism strategies, and coordinating with researchers to deliver robust, observable, and reusable frameworks. Strong Rust interest is valued as the training stack is Rust-based.

Qualifications

• Strong engineering skills with a focus on building reliable distributed systems.
• Hands-on experience in distributed model-training infrastructure and performance optimization.
• Ability to reason across the training stack and implement robust solutions.

Responsibilities

• Design and operate distributed training infrastructure across large GPU clusters.
• Improve throughput and efficiency across compute, memory, and networking.
• Develop parallelism strategies and runtime components for evolving models.

Skills

Engineering skills

Distributed training

Deep learning frameworks

Cross-functional collaboration

Problem-solving

Rust experience

Code debugging

Education

Bachelor’s degree in a relevant technical field

Tools

NCCL

PyTorch

JAX

Job description

Thinking Machines is hiring engineers to build scalable training systems that power frontier AI models. You’ll design, implement and operate distributed infrastructure across thousands of GPUs and nodes, with a focus on reliability and performance.

The role emphasizes deep understanding of training stacks, parallelism strategies, and coordinating with researchers to deliver robust, observable, and reusable frameworks. Strong Rust interest is valued as the training stack is Rust-based.