Lead Research Scientist - Post-Training
Core
Lead the end-to-end post-training pipeline for large language models, scaling training across distributed systems to produce reliable, controllable production models.
Role type
Lead Research Scientist (Post-Training & Distributed Systems)
Builds
Scalable post-training stacks including SFT, RLHF, RLAIF, and policy distillation for government and enterprise AI systems.
Domain
Artificial Intelligence / Large Language Model Post-Training / Distributed Systems
Deliverable
production ML models
Required skills
Large-scale distributed training, Supervised Fine-Tuning (SFT), Reinforcement Learning (RLHF/RLAIF), Reward modeling, Parameter-efficient fine-tuning, Fault recovery, Numerical stability debugging, Experimental design, Technical direction
Preferred skills
CUDA/Triton optimization, Long-context training, Sparse models, Asynchronous RL, Stateful agent environments, Policy consolidation
Technologies
Python, PyTorch, FSDP, DeepSpeed, Megatron-Core, vLLM, SGLang, Ray, Kubernetes, MLflow, W&B, Docker, GCP