Applied Scientist / Research Engineer

Mistral AI Mistral AI · AI Frontier · Seoul, South Korea · Solutions

Mistral AI is seeking Applied Scientists and Research Engineers to drive innovative research and collaborate with clients on complex research projects. The role involves developing SOTA models across modalities like text, image, and speech, applying them to diverse use cases, and delivering high-impact AI solutions. Responsibilities include running pre-training, post-training, and deployment on large GPU clusters, generating and curating data, developing tools for training and evaluation, and collaborating on agent and RAG pipelines. The role requires strong Python skills, PyTorch/JAX expertise, and the ability to manage research projects and client communications.

What you'd actually do

  1. Run pre-training, post-training and deploy state of the art models on clusters with thousands of GPUs. You don’t panic when you see OOM errors or when NCCL feels like not wanting to talk.
  2. Generate and curate data for pre-training and post-training, working on evaluations and making sure the model’s performance beats expectations.
  3. Develop the necessary tools and frameworks to facilitate data generation, model training, evaluation and deployment.
  4. Collaborate with cross-functional teams to tackle complex use cases using agents and RAG pipelines.
  5. Manage research projects and communications with client research teams.

Skills

Required

  • Fluent in English and Korean
  • Excellent communication skills
  • Expert with PyTorch or JAX
  • Contribute to a big codebase
  • Independent work with little guidance
  • Clean, readable, high-performance, fault-tolerant Python code

Nice to have

  • PhD / master in a relevant field
  • Experience with agents
  • Experience with multi-modality
  • Experience with robotics
  • Experience with diffusion
  • Experience with time-series
  • Contributed to a large codebase used by many
  • Track record of publications in top academic journals or conferences
  • Improving existing code by fixing typing issues, adding tests and improving CI pipelines

What the JD emphasized

  • track record of success through personal projects, professional projects or in academia
  • track record of publications in top academic journals or conferences

Other signals

  • develop SOTA models
  • apply these models across a diverse set of use cases
  • deliver high-impact AI solutions