Linjun Shou
Partner Group Applied Science Manager at Microsoft
Greater Seattle Area United States
Summary
🤩
Rockstar🎓
Top SchoolLinjun Shou is a senior applied scientist and engineering leader with 11 years at Microsoft, specializing in LLMs, document and query understanding, cross-lingual Q&A, relevance/ranking, and recommendation systems. He led Edge Machine Learning and Document Understanding teams that shipped large-scale features—such as multi-hundred-billion-scale document vectorization, long-document MRC, and QnA rolled out to 100+ languages and 200+ regions—driving Bing, Edge, and Windows copilot capabilities. A hands-on manager who still contributes code, he helped build the popular NeuronBlocks NLP toolkit and implemented core training/evaluation logic for production DNN models. He combines a strong research track record (20+ papers and benchmark initiatives like XGLUE/CodeXGLUE/GLGE) with product delivery at scale, and is skilled at turning multilingual research into deployed systems. An unexpected thread through his career is repeated success shipping language features with minimal language-specific training data, enabling rapid globalization of search and Q&A products.
11 years of coding experience
7 years of employment as a software developer
硕士, Computer Applied Technology, 硕士, Computer Applied Technology at University of Chinese Academy of Sciences学士, Mathematics, Physics & Mechanics, 学士, Mathematics, Physics & Mechanics at Shanghai University
Chinese, English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (9)
pytorch10
python10
natural-language-processing10
deeplearning-ai9
deep-learning9
artificial-intelligence8
model-compression6
question-answering5
text-classification5
Programming languages (4)
Github contributions (5)
microsoft/NeuronBlocks
Apr 2019 - Oct 2019
NLP DNN Toolkit - Building Your NLP DNN Models Like Playing Lego
Role in this project:
Back-end Developer
Back-end DeveloperContributions:4 releases, 69 commits, 53 PRs in 6 months
Contributions summary:Linjun contributed to the `LearningMachine.py` file, which suggests involvement in the core training and evaluation logic of the NLP models. Their changes include code modifications, indicating a focus on model implementation or optimization. The commits involve the modification of core model training/evaluation components within the `microsoft/neuronblocks` repository, a toolkit for NLP DNN models. This indicates a role involved in improving the training and testing of the neural network models.
knowledge-distillationbertnatural-language-processingdnntext-matching
ljshou/workspace
Nov 2014 - Feb 2016
Contributions:75 commits, 49 pushes in 1 year 4 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.