Experienced software engineer and full stack web developer interested in Large Scale Distributed Systems, Big Data, Search Engines, the Internet of Things, Music, and Machine Learning. Resume: https://resume.creddle.io/resume/7elzkjyi8cm Publications: https://scholar.google.com/citations?user=8Pc5MiUAAAAJ&hl=en Email: kevd1337@gmail.com Phone: 214.435.4949

Software Development Engineer II @ Worked for the Matching team which develops and maintains the service responsible for matching millions of new product offers per day against Amazon's existing catalog: ■ Developed image feature extractors for use in machine learning based item de-duplication efforts for the Amazon catalog ■ Led the re-architecture and resource planning of Amazon's external identifier cross reference system from a massive RDBMS to a NoSQL data store which allowed for a live migration while continuing to serve high TPS traffic without data loss or availability outage ■ Leveraged information extraction algorithms applied to attributes in product descriptions to more accurately match seller submissions without barcodes against existing products in the catalog leading to a reduction in duplicate items in Japan ■ Implemented a web tool used daily by hundreds of seller support agents worldwide to correct mismatched seller product listings in the Amazon catalog. Drove down related merchant contacts by 43% and reduced the time to resolve such issues into a few mouse clicks ■ Built tools for auditors to measure quality of catalog impacting decisions made by the item authority's product matching algorithms, leading to insights and policy changes that reduced duplicate creation for non-Novelty items by an estimated 58% in the US ■ Enabled the launch of 30 new product categories in multiple countries into Amazon's catalog by collaborating with the business teams to develop appropriate product validation logic and implementing matching criteria ■ Drove engineering excellence by instituting code review guidelines, establishing team development standards, bar raising deployments for Amazon Marketplace team, leading software design reviews and leading sprint planning & retrospective ■ Trained and mentored a handful of new employees in Seattle and Hyderabad, India ■ Active in recruiting having conducted over 125 technical interviews From January 2013 to Present (3 years) Greater Seattle AreaSoftware Development Engineer I @ Developed and maintained Herd, a distributed workflow platform at the heart of Amazon's e-Commerce platform which many teams have come to rely on for building scalable applications: ■ Drove the re-factoring of the platform and associated back-fills and customer communication to enable multi-tenancy and safe client isolation allowing us to triple the number of clients on the platform within the timescale of months without sacrificing availability ■ Conducted office hours to consult with clients on application design From September 2011 to January 2013 (1 year 5 months) Greater Seattle AreaResearch Assistant @ ■ Assessed the impact of various methods of instruction on ESL students through a web based intelligent tutoring system called REAP by developing new NLP driven features, collaborating with instructors on curriculum, and conducting field research with students at the University of Pittsburgh ■ Conducted self directed research in the fields of natural language processing, information retrieval, social media analysis and machine learning. ■ Resulted in 8 publications in under 2 years in venues such as SIGIR, Interspeech SLATE, ACL-HLT, AIED, MLSLP, FLAIRS From January 2010 to August 2011 (1 year 8 months) Naval Research Enterprise Intern @ Developed text classifiers using naive bayes, rocchio, k-nearest neighbor, and support vector machine algorithms for use on military chat posts. Results published in ICMLA 2009 From May 2009 to August 2009 (4 months) Research Assistant, Teaching Assistant @ Vision Learning Mining Research Lab - Research Assistant [ March 2008 – May 2009 ] ■ Developed ranking functions for recommendation systems and created a boosting framework to improve their performance on ranking user movie preferences. Results published in KAIS Department of Physics - Research Assistant [ January 2009 – May 2009 ] ■ Collaborated with NASA scientists to improve existing solar supergranule feature rotation simulation code base to match latest data collected from NASA's SOHO satellite's MDI instrument. Results published in the Astrophysical Journal Department of Computer Science - Teaching Assistant [ August 2008 – May 2009 ] ■ Teaching assistant for Object-Oriented Software Engineering, Graduate Software Engineering Automation & Robotics Research Institute - Research Assistant [ October 2007 – May 2008 ] ■ Produced physics based 3D simulations of airborne and land based micro-scale robots with Microsoft Robotics Studio and 3ds Max. From October 2007 to May 2009 (1 year 8 months) Summer Undergraduate Research Fellow @ Developed an AJAX/GWT interface for a software testing tool called FireEye as part of a research project involving automated combinatorial software testing. From May 2007 to August 2007 (4 months)

Master's Degree, Computer Science - Language Technologies @ Carnegie Mellon University From 2009 to 2011 Bachelor's Degree, Software Engineering & Physics @ The University of Texas at Arlington From 2004 to 2009 Kevin Rosa is skilled in: Machine Learning, Natural Language Processing, Java, Information Retrieval, Python, Algorithms, Software Engineering, Hadoop, Scalability, Amazon Web Services (AWS), Big Data, Apache Spark, Distributed Systems, Software Architectural Design, Full Stack