Hands-on applied research role developing post-training algorithms and open-source tooling for NLP. Prototype and optimize deep learning models with PyTorch or JAX, run large-scale GPU experiments, contribute to OSS, and publish empirical results.