Latest AI JobsBrowse Jobs
Thinking Machines logo

Research Engineer, Infrastructure, Numerics

Thinking MachinesSan FranciscoResearch Infrastructure (ML Infrastructure and Training Stack)
Apply on company site
Salary
$350K - $475K USD
Job type
Full-time
Work mode
Remote
Posted
4 Aug 2026
Status
Open · checked 7d ago

About this role

Role Overview

The Research Engineer, Infrastructure, Numerics position at Thinking Machines is a full-time role based in San Francisco, California, with remote flexibility available. This opportunity is well-suited for a skilled builder who enjoys working at the intersection of systems engineering and research to support large-scale model training.

Key Responsibilities

Architect and refine distributed training systems for large language models, ensuring speed, stability, and reproducibility across multi-node and multi-GPU environments.
Test and integrate low-precision numerics like BF16, MXFP8, and NVFP4 to maximize efficiency while preserving model quality.
Build specialized communication primitives and kernels that leverage hardware capabilities for mixed-precision and low-precision calculations.
Partner alongside research groups to shape model architectures and training protocols tailored to emerging numeric formats and stability limits.
Test and measure scaling methods including tensor, data, and pipeline parallelism paired with quantized communication and precision-adaptive computing.
Shape internal monitoring and orchestration platforms to support smooth and reproducible execution for thousands of distributed experiments.
Share insights via open-source tools, internal documentation, and technical reports to push the boundaries of scalable artificial intelligence infrastructure.

Requirements

Hold a bachelor's degree or equivalent practical experience in computer science, machine learning, electrical engineering, statistics, physics, robotics, or a related field.
Bring a solid working knowledge of deep learning tools like PyTorch and JAX alongside their internal system designs.
Excel in team-oriented settings working alongside various cross-functional experts and subject matter specialists.
Demonstrate an active bias for action, taking the initiative to span different tech stacks and teams to deliver results.
Exhibit robust engineering capabilities with the aptitude to write clean, maintainable code and troubleshoot intricate codebases dealing with distributed systems, low-precision arithmetic, and floating-point numerics.

Nice to Have

Prior background with distributed frameworks such as Megatron-LM, DeepSpeed, or PyTorch/XLA.
Hands-on practice deploying INT8, FP8, or block-floating point (MX) data formats alongside an understanding of their numerical impacts.
Past contributions directed toward open-source deep learning platforms including XLA, DeepSpeed, or PyTorch.
Documented publications, patents, or completed projects focusing on communication-efficient training, numerical optimization, or large model systems.
Direct involvement in training and maintaining large-scale artificial intelligence models.
A proven history of boosting research efficiency through thoughtful infrastructure planning and workflow enhancements.

Skills and Tools

PyTorch
JAX
DeepSpeed
Megatron-LM
PyTorch/XLA
Distributed Systems
Low-Precision Arithmetic
Floating-Point Numerics

Salary and Benefits

Annual salary range: $350,000 to $475,000 USD, determined by background, skills, and experience.
Comprehensive medical, dental, and vision insurance coverage.
Flexible unlimited paid time off (PTO).
Paid parental leave programs.
Relocation assistance provided when necessary.
Visa sponsorship support available for qualified candidates.

About the Company

Thinking Machines works to create artificial intelligence designed to expand human will and decision-making capabilities. The organization trains cutting-edge models using Inkling, builds Tinker to help users personalize models, and crafts communication interfaces for humans and AI.

How to Apply

Interested candidates should visit the official Thinking Machines careers page to submit an application for this opportunity.

Skills and tech stack

PyTorch

Questions candidates ask

What is the salary range for the Research Engineer, Infrastructure, Numerics role?

The expected annual salary ranges from $350,000 to $475,000 USD, depending on experience, skills, and background.

Where is this position located?

The job is based in San Francisco, California, with remote flexibility.

Does Thinking Machines offer visa sponsorship?

Yes, Thinking Machines sponsors visas and works closely with candidates who are the right fit.

What benefits are included with employment?

The company provides health, dental, and vision benefits, unlimited paid time off, paid parental leave, and relocation support if needed.

Is this an evergreen job posting?

Yes, this is an evergreen role kept open continuously to gather expressions of interest, and applications are reviewed on an ongoing basis as new opportunities arise.

Ready to apply?

The application is on the official Thinking Machines careers page.

Apply on company site