Back
Tier-1 brand, mid-level (5+ years) role, and metro location drive high competition.
Role requires specialized LLM training and research expertise, so cross-industry transferability is low.
Explicit 5+ years plus mandatory LLM training, PyTorch, and agentic research requirements make filters strict.
Login to See Your Match Score
Create a free account or log in to unlock your CV match score across:
Lead mid-training and post-training research for LLMs including continued pretraining, SFT, preference optimization, and RL with focus on how training decisions affect agent behavior.
Prototype and research novel agentic architectures and algorithms in planning, reasoning, memory, skills, tool use, retrieval, and multi-agent collaboration advancing beyond current methods.
Design and implement research frameworks and evaluation methodologies enabling reproducible experiments, trajectory analysis, and rigorous benchmarking of agent capabilities and failure modes.
5+ years of experience in machine learning, deep learning, AI research, or related field with demonstrated applied research experience.
Hands-on experience with LLM training and post-training workflows including continued pretraining, SFT, preference optimization, or RL.
Strong Python and advanced PyTorch skills, including modifying models and training pipelines for experimentation.
Experience designing evaluation methodologies such as benchmark design, trajectory-based evaluation, LLM-as-a-Judge, and human evaluation.
Ability to independently define research problems, formulate hypotheses, and drive projects to validated prototypes with measurable impact.
Deep practical knowledge in agentic AI domains covering planning, reasoning, memory, tool use, retrieval, and context/knowledge engineering.
Demonstrated track record of communicating complex research findings effectively to technical teams and executive stakeholders.