CV

Dustin Rigg Hillard
VP of Engineering, Context Relevant
Seattle, WA 98117
dustin.rigg.hillard [on] gmail.com

Expertise:

Machine learning at scale.

Education:

University of Washington, Department of Electrical Engineering

    • Bachelor of Science Degree, 2002, cum laude, with College Honors, Music Minor
    • Master of Science Degree, 2004
    • Doctor of Philosophy, 2008
      • Dissertation: Automatic Sentence Structure Annotation for Spoken Language Processing
      • Advisor: Professor Mari Ostendorf

Employment:

2012 – current            Context Relevant, Seattle, WA

VP of Engineering

Responsible for delivering machine learning at scale.

2010 – 2012            Microsoft, Redmond, WA

Senior Scientist, Bing

Improving spoken language understanding

  • Contributed novel named entity recognition approach based on mining web click logs
  • Helped establish crowd sourcing approach for bootstrapping models for new tasks
  • Improved multiple components of a spoken dialog system
  • Published 4 conference papers, submitted 1 patent

2008 – 2010            Yahoo! Inc., Santa Clara, CA

Scientist, Sponsored Search Applied Sciences

Responsible for delivering new approaches to improve revenue and quality:

  • Contributed five novel algorithmic improvements to ad retrieval that resulted in significant increase in revenue per search (RPS)
  • Built Yahoo’s first model of query-ad relevance for sponsored search, which filtered low quality ads and improved ranking
  • Improved click and relevance models with query segmentation features, providing better accuracy for tail queries
  • Frequent owner of online experimentation and evaluation ‘buckets’
  • Strong collaborations across research, science and engineering teams
  • Large scale learning from user logs using Hadoop/Pig/MapReduce
  • Implemented all contributions in production quality C++
  • Published 5 conference papers, submitted 2 patents

2002–2008            University of Washington, Seattle, WA

Research Assistant

Research involving supervised and unsupervised learning of sub-sentence prosodic boundaries and their use for improving natural language processing. Also developed new approach for improved confidence estimation in speech recognition output. Earlier work focused on improving automatic sentence boundary detection using multiple speech recognition hypotheses.

2006                        RWTH, Aachen, Germany

Visiting Researcher (NSF Sponsored)

Research in automatically predicting commas for Mandarin and Arabic to improve POS tagging, named entity tagging, and machine translation. Also developed novel techniques for improved ASR system combination.

2001-2002            University of Washington, Seattle, WA

Undergraduate Research Assistant

Research funded by NSF REU Award to determine regions of agreement and disagreement in meetings.  Designed and implemented a system that processed meeting transcripts to identify regions of consensus and dissent.

1999-2001            Siemens Ultrasound Group, Issaquah, WA

Hardware Engineering Intern

Designed, tested, and implemented C++ and C language code for detection and correction of hardware failures.

Awards:

      • Yahoo! You Rock Team Award, 2009, 2010
      • Yahoo! Superstar Team Award, 2008, 2010
      • MIT Lincoln Labs Research Fellowship, 2006-2007
      • NSF Doctoral Dissertation Enhancement Project Grant, 2006
      • Outstanding Undergraduate Research Assistant Award, 2002
      • Washington Scholar, (4 year full tuition scholarship) 1998
      • Valedictorian, (in a class of 450) 1998
      • Eagle Scout

Consulting:

2006 –2007            Program on Networked Governance, Kennedy School, Harvard

Visiting Research Fellow

Research using machine learning with active learning to automatically categorize congressional legislation and congressional websites to enable Political Science analysis for the Connecting to Congress project.

Publications:

See my publications page