Senior ML Research Scientist, Jockey Core
Twelve LabsVideo Understanding company
Seoul, South KoreaSenior
Data & AI
About the role
TL;DR
Lead model efficiency research for a critical reasoning LLM.
- •This role leads model-efficiency and post-training research for Jockey Core, a critical reasoning LLM, aiming to make high-quality models production-efficient without sacrificing performance.
- •You will drive research in model compression and efficiency, design rigorous evaluations, and explore post-training techniques.
- •Key Responsibilities Drive model compression and efficiency research (pruning, quantization, distillation).
- •Design rigorous evaluations on agent's real reasoning and tool-calling behavior.
- •Explore post-training (SFT/RL) to preserve or improve agentic tool-use.
- •Work closely with serving engineers to translate efficiency gains into cost and latency wins.
- •Requirements Strong LLM research experience (post-training, model compression, distillation, efficient inference).
- •Track record of independently driving research from ideation to execution.
- •Strong experimental judgment (eval design, rigorous ablations, clear empirical conclusions).
- •Strong proficiency in Python and PyTorch.
- •Ability to communicate and collaborate closely with researchers and engineers.
Required skills
PythonPyTorchLLMsComputer VisionNLP
Nice-to-have skills
TensorFlowscikit-learnKeras
Domain expertise
aideveloper-tools
Benefits & perks
Global Team, Hybrid work, Stock options, Unlimited LLM tokens, Self-development budget, English language training, Global buddy program, Night and weekend taxi reimbursement, Meal & transportation allowance, Snack bar, Dinner provided, Annual health check-up, Family health insurance, Maternity/Paternity leave, Childbirth support, Smoking cessation support, Yearly Holiday Break
Tech stack
PythonPyTorchLLMsComputer VisionTensorFlowKerasscikit-learn