Peter Izsak
Applied research on language models: efficiency, retrieval, and evaluation.
I’m an applied research scientist at Microsoft, working on NLP for healthcare. I currently work on agentic clinical systems, retrieval-augmented generation, and evaluation of generative and agentic AI.
Before Microsoft I was a research scientist and team lead at Intel Labs, working on deep learning for NLP. I hold an M.Sc. and B.Sc. from the Technion — Israel Institute of Technology.
My research covers efficient language-model training, quantization, retrieval-augmented generation, long-context evaluation, and knowledge distillation, with publications at EMNLP, NeurIPS, COLING, and ACM SIGIR. Earlier work covered information retrieval and distributed systems.
