Jobs

As a Research Engineer, Post-Training, you will be a critical bridge between raw model capability and real-world utility for end-to-end software agents like Devin. This role involves shaping how AI agents learn by iterating on training recipes, evaluations, and alignment methods, blending deep research with hands-on engineering. You will develop post-training recipes, design robust evaluations, and apply advanced techniques like RLHF and RLAIF to ensure agents reason, act, and collaborate effectively. We are looking for candidates with a strong track record in advancing ML systems through post-training and alignment, deep ML fundamentals, and experience with large-scale distributed training.