SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Mistral is seeking an AI Scientist to research and develop novel methods that advance the capabilities of large language models. You will join a team of engineers and scientists focused on building tooling and infrastructure to train, evaluate, and analyze AI models at scale, working across multiple use cases and modalities including reasoning, code generation, text, image, and speech processing.
In this role, you will research and develop innovative methods to push the frontier of large language models, build tooling and infrastructure enabling large-scale training, evaluation, and analysis of AI models, and collaborate cross-functionally with scientists, engineers, and product teams to deploy AI systems with real-world impact. You will design and implement solutions addressing complex challenges in AI model development and deployment, contribute to the open-source community by releasing cutting-edge models and tools, and drive projects forward with autonomy while working effectively in a team environment.
You should have a strong publication record in machine learning or natural language processing, high proficiency as a software engineer in languages like Python, Rust, Go, or Java, and hands-on experience with AI frameworks such as PyTorch or JAX, or distributed systems like Ray or Kubernetes. Experience training large transformer models in a distributed manner is essential, along with familiarity across the full MLOps stack including fine-tuning, evaluation, and deployment. Experience with audio or speech processing, including input/output and NLP, is valued. Mistral operates as a dynamic, collaborative team distributed across Europe, North America, Asia, and the Middle East, passionate about AI's potential to transform society and committed to driving innovation.
