Jobs with visa sponsorship in United States

About the role

We’re hiring a  research engineer focused on developer experience  to build Tinker while using Tinker — working hands-on with real users and turning their challenges into product improvements. You’ll write and update cookbook recipes, add library features,  prototype integrations, and ensure users can customize models smoothly. This role is a bridge between Tinker users and our research and infrastructure teams. You’ll surface user patterns to guide product and infrastructure priorities, and share what you learn through blog posts, demos, and code examples. This is a technical-first role with a major user-facing component. It’s ideal for someone who’s happiest in a code editor or Jupyter notebook, but who also enjoys talking to creative people. Tinker is flexible by design to empower novel and ambitious use cases for model fine-tuning — your job is to help them succeed.

What you'll bring

• Enabling user success in building on top of an advanced platform is a skill that can be developed through many paths.
• Experienced with fine-tuning large language models (LLMs) with supervised and reinforcement learning, and how to tune the hyperparameters of these algorithms.
• Experienced with developing software libraries, especially open-source libraries.
• Are thoughtful about developer experience, with clear ideas around developer ergonomics, onboarding flow, and sharp edges.
• Are passionate about engaging with open-source communities, whether by contributing code, answering questions, or sharing learnings.
• Have written public or internal code examples, tutorials, or papers that helped others understand and apply technical concepts.
• Are comfortable reasoning about model training, fine-tuning, and inference, and can credibly discuss tradeoffs between different approaches.

Benefits