Youngsuk Park Email and Phone Number

I am a senior applied scientist & team leader of 10+ PhD scientists and research interns at AWS AI Research and Education (AIRE). I establish a science roadmap of efficient training and inference for foundation models. My team delivers algorithmic solutions ranging from distributed & robust training, modern compression, to foundation modeling via leveraging modeling, theoretical analysis and system knowledge effectively. Our solutions are ultimately system-hardware compatible and even best utilization of unique hardware features to reflect in new algorithms, bringing real world impacts to internal (e.g., Amazon Search and AGI) and external customers (via Amazon Bedrock, Sagemaker, Antrophic). I have actively been publishing 30+ papers in the area of optimization, LLM, RL, time series at top venues (ICLR, ICML, NeurIPS, AISTATS, KDD) where recent papers focus on efficient training and inference, e.g., robust training with stochastic rounding, low-precision training, structured sparsity, inference-based optimization. I delivered multiple tutorials and workshops, e.g., LLM training on AWS Trn, Inference optimization for LLM on AWS Trn, Modern time series modelling, etc. I actively serve as PC/AC in ML communities. See https://youngsuk0723.github.io/ .My team is constantly hiring researchers. Please reach me out for the opportunities if interested.