I currently work in the application of machine learning to the problem of matching employers and employees. We use various data mining and information retrieval techniques to overcome the limitations of our sparse input data (member profiles and job descriptions). I honed my data and programming chops as as physicist working in large nuclear and astrophysics collaborations. We developed algorithms to process 100 MB/s worth of data in search of very rare signals.

Sr. Manager, Engineering @ I manage a team of scientists and engineers that builds recommendation systems for LinkedIn Talent Solutions products. We focus primarily on matching members and jobs through profile based recommendations (Jobs You May Be Interested In) or collaborative filtering. Our team has strength in both fundamental machine learning and domain knowledge in Person-Job fit. From March 2014 to Present (1 year 10 months) Chief Scientist @ Bright.com (acquired by LinkedIn in Feb., 2014) built algorithmic tools to match candidates and employers. I lead a team of data scientists and engineers who developed the Bright Score, a data driven relevancy algorithm that helps employers immediately identify the most suitable candidates from a pile of resumes and helps employees immediately find the best available job openings. From November 2011 to March 2013 (1 year 5 months) Director of Search @ Cloudant built a fault-tolerant framework for real-time search and analytics on massive datasets. I lead the initial efforts to build a scalable search product on top of CouchDb. From January 2010 to November 2011 (1 year 11 months) Founder and Developer @ Developed a new meta-search engine that utilized the real-time web to improve the timeliness and pertinence of web, video, and image search. From May 2009 to November 2011 (2 years 7 months) Chief Scientist @ Surf Canyon developed implicit relevance feedback technology for web search. User clickstream and behavioral data are used to improve the search results page as the user searches. I developed the implicit feedback algorithm used for both organic search results and sponsored links. I also invented an intelligent query reformulation technology that substantially increases user engagement with the Surf Canyon search page. From August 2007 to May 2009 (1 year 10 months) Research Scientist @ Neutrino Astrophyiscs Research using the IceCube Neutrino Observatory at the South Pole Station From August 2003 to August 2007 (4 years 1 month) Staff Scientist @ Experimental Nuclear Physics Research. Worked on Solenoidal Tracker at RHIC (STAR) experiment at Brookhaven National Lab. From December 1997 to August 2003 (5 years 9 months) Visitng Researcher at RHIC @ From 1997 to 2003 (6 years) Grad Student Researcher @ Thesis work on NA44 experiment at CERN (Geneva, Switzerland) From 1994 to 1997 (3 years) Visiting Researcher @ From 1994 to 1997 (3 years)

B.S., Physics, Canadian Studies @ Duke University From 1990 to 1994 Ph.D., Experimental Nuclear Physics @ The Ohio State University From 1994 to 1997 David Hardtke is skilled in: Algorithms, Analytics, Java, Python, Linux, Data Analysis, Statistics, Data Mining, Distributed Systems, Science, Machine Learning, Perl, Programming, LaTeX, Start-ups