Pi-Chuan Chang Email & Phone Number

Mountain View, Ca, Us

2017Q2-Present: Genomics in Google Health(Tech Lead / Manager for DeepVariant and DeepConsensus) - Team blog: https://goo.gl/deepvariant - DeepVariant: https://github.com/google/deepvariant - Nature Biotechnology paper : https://rdcu.be/7Dhl - DeepConsensus: https://github.com/google/deepconsensus - Nature Biotechnology paper: https://rdcu.be/cUMIx 2016 Google Research, Summarization and Sentiment(Tech Lead for Text Summarization)

Sunnyvale, Ca, Us

Natural Language Processing and Content Relevance @ LinkedIn.◦ Related Skills:Built the backend mapping of the Related Skills module on the topic page like https://www.linkedin.com/topic/machine-learning◦ Article Features workflow:Built the Article Features data workflow: extracting features (such as skills, topics, etc) from articles, and ETL for all internal client teams to build various recommendation pipelines◦ Query Tagging:Revamped the query tagger used in search - starting from data collection, evaluation, building a discriminative sequence model (CRF) and feature engineering to utilize LinkedIn specific taxonomy resources, and all the way to plugging the new tagger into the production system for A/B testing.

◦ Cross-functional collaboration: I worked with Educators, Ops and Product/Design/Engineering team. It has helped to iterate faster and improve product features in a user-driven manner because I was able to work well with everyone and also functioned as different roles as needed (such as UXR, a teacher, an engineer, etc).◦ Product & Engineer: I was the lead frontend engineer of several teachers-facing products, including the Chrome extension for editing students' playlists, and the inline assessment tool for documenting and assessing student's progress both on academic and social emotional learning.◦ Research: - led Social Emotional Learning (SEL) working group with educators to build learning objectives, activities to effectively teach SEL in AltSchool classrooms. - researched and prototyped ways of demonstrating more academic rigor, such as including an formative assessment item bank in students' playlists, and integrating Khan Academy progress data into our learning cycle.

Mountain View, Ca, Us

Semantic parsing in Conversational Speech (2013/03 -- 2013/10) ◦ scaled up the training component for the parser; built the distributed training system which trains the model on MapReduce and allows scaling for much larger training sets and much shorter training timePersonalization across Google properties (2011/09 -- 2013/03) ◦ Quality: built the initial user model and experiment framework, collected various sources of data, designed features, designed metrics, monitored and improved quality of the model ◦ Client collaboration: worked directly with product teams, such as YouTube and Shopping, to improve personalization.Semantic Parsing in Google Research (2009/10 -- 2011/09) ◦ worked on a research prototype system that parses user queries as well as Web documents into semantic representations ◦ published two conferences papers based on our work: 1. "Deterministic Statistical Mapping of Sentences to Underspecified Semantics" in IWCS 2011. 2. "Uptraining for Accurate Deterministic Question Parsing" in EMNLP 2010. ◦ worked with different search feature teams (such as search result snippets and question-answering) to use semantic parsing to improve their features ◦ launched a coverage improvement for question-answering in search

Stanford, Ca, Us

◦ Research area: Natural Language Processing and Machine Learning specifically on : Statistical Machine Translation (specifically, Chinese to English), Chinese word segmentation (CRF sequence modeling), Chinese Grammatical Relations in the Stanford Parser.◦ TAed Natural Language Processing (CS224N)

Mountain View, Ca, Us

Worked on search quality in the Image Search group using textual information around images

Redmond, Washington, Us

Worked in the Natural Language Processing group at Microsoft Research, Redmond.The work done in the summer was then published in ACL 2007:"A Discriminative Syntactic Word Order Model for Machine Translation"Pi-Chuan Chang, Kristina Toutanova. In The Annual Meeting of the Association for Computational Linguistics, June 2007.

Doctor Of Philosophy (Ph.D.), Computer Science

Bachelor Of Science, Computer Science