Applied AI Researcher, Post-Training at Distyl AI

The position is hybrid.

This information is not specified in the job description.

Required skills include deep understanding of post-training techniques like supervised fine-tuning, preference optimization (RLHF/DPO), LoRA/PEFT, and instruction-tuning; experience adapting frontier models through data curation, reward modeling, or continual pretraining; and expertise in building compound AI systems, agentic collaboration, ensembling, ReAct, and graph-of-thoughts.

Distyl AI features a fast-growing, innovative environment pushing AI boundaries in enterprise settings, with researchers who think creatively, operate AI-natively, and redefine software use rather than following traditional paths.

Strong candidates have strong research track records, experience adapting frontier models to real-world needs, and expertise building intelligent systems with models using techniques like agentic collaboration, rather than just training models.