Research Engineer, Forge

Sourced from careers.redpoint.com

Mistral AI is seeking a Research Engineer for its Forge team to bridge the gap between research experimentation and production constraints. You will build end-to-end workflows for model adaptation, post-training (SFT/RL/distillation), and evaluation, turning customer requirements into reliable, scalable AI systems. This role involves developing tools for synthetic data generation, debugging large-scale ML infrastructure, and optimizing distributed training pipelines.

• Build and improve post-training and evaluation workflows (CPT/SFT/RL/distillation) to create repeatable recipes.
• Develop tools and pipelines for synthetic data generation, curation, training, evaluation, and deployment.
• Debug and harden large-scale ML systems including distributed training, scheduling, checkpointing, and observability.
• Improve the Forge codebase with clear APIs, tests, documentation, and maintainable abstractions.
• Push the frontier of RL training stacks at frontier-model scale.

• Strong Python engineering skills with experience in large codebases, testing, CI, and operational ownership.
• Hands-on experience with PyTorch, JAX, or similar deep learning frameworks.
• Experience with LLM training or post-training techniques such as fine-tuning, RL, distillation, and evaluation.