4+ year's hands-on experience in machine learning algorithm design as well as large scale data pipeline architect. Expertise in both theory and practice of machine learning. Extensive research experience in human activity recognition, time series analysis, data collection, human behavior modeling and personalization (with deep practical understanding of how and when it should happen); knowledgeable in Natural Language Processing. I am also a software engineer experienced in large scale data infrastructures. With complete understanding of full stack and proven industry record in delivering firmware, mobile and data pipeline. Have working experience with Hadoop MapReduce & Spark with Petabyte scale of data. Able to self-architect backend infrastructure for data driven products. Solid understanding in object oriented programming and experienced in Java/C/C++/Scala. Enjoy the art of programming. Fast learner and comfortable with new technologies. Specialties: Algorithm, Machine Learning & Data Modeling, Hadoop, Spark. Programming Language: Java, Scala, C/C++, Python, Ruby on Rails

Software Engineer @ Building salable data infrastructures to power up Apple Map and related products. From July 2015 to Present (6 months) San Francisco Bay AreaData Scientist @ Participate in the full development cycle of Hello Sense and deliver its core infrastructure (Sensing/Processing/Data storage/Alarm/Timing system). Develop the ultra low power firmware of Sense Pill. Including design its ultra low-power feature extraction algorithm, as well as the data transmission scheme, which makes it possible for 8+ months battery live on a coin cell battery. Also initiated the algorithm researching for sleep event detection using pill data. From February 2014 to July 2015 (1 year 6 months) San Francisco Bay AreaSenior Research Engineer @ Architect and build the sensing platform for large scale crowd source mobile context data collection. The platform features provider recruitment (Integrates with oDesk), large scale sensor data processing and visualization. I also test out different algorithms for context-aware applications based on the data collected from the sensing platform. From May 2013 to February 2014 (10 months) San Francisco Bay AreaResearcher, PhD @ Focus on using mobile sensing technique for Personalized Activity Recognition Human Behavior Modeling. Designed and prototyped CMU MobiSens, an App for super lightweight on-device sensing and activity recognition; and Lifelogger (http://www.lifelogger.info): A versatile platform for real-world user behavior modeling which supports unsupervised activity segmentation. The platform is also a testing bed for multi-modality activity recognition algorithm. From September 2011 to October 2013 (2 years 2 months) San Francisco Bay AreaResearch Intern @ Work with the Personalized Relevancy, Ranking, and Recommendation (PR3) Team at Nokia Research Palo Alto. Design tools to support the collection and visualization of mobile context data. From May 2012 to August 2012 (4 months) Project Owner and Technical Lead @ MSNPSharp (http://code.google.com/p/msnp-sharp/) is a .net library that implements the MSN protocol (A C# implementation of MSN protocol). It's users includes Mxit.com, Monkey Messenger, GamersBook. Contribution: - Reverse Engineered MSN Protocol Version 11/15/16/18/21 including the binary protocol for P2P data transfer and the web service schema introduced in MSNP15. - Re-implemented the MSN Protocol stack of above version using C#. - Designed and implemented an easy-to-use, asynchronous API library for third party MSN developers to leverage Windows Live IM service. TCP/IP and Socket Programming, C#, Webservice development and Reverse Engineering. MSN Protocol analysis. From March 2007 to January 2012 (4 years 11 months) Software Engineer @ Worked with QQMail/ReaderSpace development team and took responsibility of developing the distributed data serves over 100 million active users in Tencent’s service. Programming language and Environment: C/C++, Shell Programming, Linux/Unix From March 2008 to January 2009 (11 months)

PhD Student, Electrical and Computer Engineering @ Carnegie Mellon University From 2011 to 2013 Master, Software Engineering, 3.93/4.0 @ Carnegie Mellon University From 2010 to 2011 Master, Information Technology @ The Hong Kong University of Science and Technology From 2009 to 2010 Bachelor, Software Engineering @ South China University of Technology From 2005 to 2009 Pang Wu is skilled in: C#, C++, Java, jQuery, Ruby on Rails, Sockets, Machine Learning, Android, HTML + CSS, Linux, Python, JavaScript, Unix Shell Scripting, SQL, Activity Recognition System