Over the last few months, Midjourney has rapidly developed a computationally intensive web and Discord-based service that has scaled quickly to provide state-of-the-art image generation AI to millions of users on a daily basis. We’re just getting started, both on our growth trajectory as well as on the generative model capabilities we want to deliver. We’re going to be pushing the boundaries of what’s possible as we develop and refine our collection of ML models.

Since we’re 100% remote, you’ll need to be able to be productive in a relatively unstructured, self-directed environment, and you’ll need to have excellent written and verbal communication skills to help keep everyone in sync. You’ll also need to have excellent time management skills.

Join our talented team of AI researchers to help us solve hard problems for our growing user base!

Responsibilities:

  • Design, test, and tweak the core ML systems that enable image generation, modification, and analysis.
  • Split the awesome, interesting, fun work with the team
  • Split the unglamorous, annoying, dirty work with the team
  • Be both an expert and a team player

Requirements:

  • 5+ years of deep learning experience in a commercial or research environment
  • Experience being the primary researcher responsible for the development, testing, and productization of complex ML models
  • Experience with training large machine learning models
  • Strong background in mathematics (especially linear algebra, differential equations)
  • Familiarity with state-of-the-art generative image models, including diffusion models
  • Experience with JAX or training on TPUs is a strong bonus