GrepJob
Thinking Machines Lab

Software Engineer, Developer Productivity, AI Tools

Thinking Machines Lab
Apply
5 days ago
San Francisco, CA, USAMid Level
H1B Sponsor

Base Salary

$350k - $475k/yr

Responsibilities

  • Enable researchers and engineers to use AI to improve coding productivity without compromising code quality.
  • Standardize, configure, harden, maintain, and integrate AI coding tools including Claude Code, Cursor, and Codex.
  • Build secure, reproducible agent sandboxes for remote development and CI testing.
  • Set up golden-path development environments and guardrails for secrets and PII.
  • Help individual contributors develop personalized AI-enabled workflows.
  • Track AI coding tool usage, reliability, and cost.
  • Partner with platform, security, product, and research engineers to improve internal software development workflows.

Requirements

  • Bachelor’s degree or equivalent industry experience in computer science, engineering, or a similar field.
  • Experience developing productivity tools and best practices for large codebases.
  • Ability to communicate clearly and work with researchers to build and manage varied internal tools.
  • Hands-on experience with container platforms such as Docker or Kubernetes is preferred.
  • Experience with modern CI tools such as GitHub Actions or Buildkite and package management tools such as uv is preferred.
  • Practical experience with AI coding tools and model APIs, including OSS tools through vLLM, SGLang, or TGI, is preferred.
  • Solid Linux and networking fundamentals, with comfort in secrets management and safe egress, are preferred.
  • Proficiency in systems programming languages such as Rust and scripting languages such as Python is preferred.

Benefits

  • Generous health, dental, and vision benefits.
  • Unlimited paid time off.
  • Paid parental leave.
  • Relocation support as needed.
  • Visa sponsorship is available.
  • The role is based in San Francisco, California.
  • This is an evergreen role reviewed on an ongoing basis rather than a single immediate opening.

Tech Stack

BuildkiteDockerGitHub ActionsKubernetesLinuxPythonRust

Categories