Home Resume Media Projects Contact
 
     
 

RESUME

  Name Anirudh Koul
  Past Job Profile Research Engineer at Yahoo Inc (4 Years)
  Current University Carnegie Mellon University (School Of Computer Science)
  Degree Masters in Very Large Information Systems (MS VLIS)
  GPA A (4.08 / 4)
  Expected graduation Date June 2012
  Country of Citizenship Canada
  Email Address mylastname@cs.dal.ca (Sorry spam bots)
  Latest Resume View Anirudh Koul's profile on LinkedIn

left Skill Set right
  • Technical : Machine Learning on Large Scale Data, Natural Language Processing, Scalability, Information retrieval, Recommender Systems, Web Mining, Sentiment Analysis, Fast Prototyping
  • Non Technical : Project Management, Intra-preneurship, Public Speaking

Technical Achievements/Awards
  • Innovator of the Year award from National Geographic Channel for the invention of Non Redundant Data Compression Technique. (December 2004)
  • Young Innovators Award 2004 sponsored by IBM and CII-DST (Confederation of Indian Industries and Department of Science & Technology, GOI), India. (February 2005)
  • Best Paper Award for paper titled Determination of Sequences in Prime Numbers, National Conference on Bioinformatics Computing, India. (March 2005)
  • Finalist at Infosys EducationWorld Young Achiever Award (July 2005). Featured in the cover story of EducationWorld Magazine. (August 2005)
  • First featured user on Yahoo Answers (January 2006). Also featured in New York Times (February 2006) and Chronicle Herald, Canada. (April 2006)
  • Undergraduate Student Research Award from Natural Sciences and Engineering Research Council (NSERC), Canada. Topic of research ‘Automatic Network Traffic Classification for Peer-to-Peer Traffic’. (September 2006)
  • Honorable Mention in Computing Research Association (CRA), 2007 Outstanding Undergraduate Award sponsored by Microsoft Research. (December 2006)
  • At Yahoo: ‘You Rock’ award (for building applications that eliminate 70% of spam before reaching Customer Care) (October 2009). ‘You Rock’ team award for ÔOutstanding contribution for a newly formed teamÕ (September 2010). Also, declared finalist for Q2 Yahoo European Award of Recognition. (September 2008)
  • At Carnegie Mellon: Yahoo Hackathon 2010, Facebook Hackathon 2011, Yahoo Hack Pitch 2011, Google Demo (For NLP class) - Most Innovative Question Answering system
left Programming Skills right
  • Proficient in PHP, Java, C/C++, SQL. Frequent on-demand experience with Perl, Python, R, Shell, Pro*C, HTML, JavaScript, PL/SQL
  • Toolkits and Packages: Hadoop, Mahout, Mallet, Weka, Stanford NLP, OpenNLP, NLTK, Lucene

left Work Experience right

  • Yahoo Inc.: Research Engineer (May 2007 – June 2011)
    Reported to the Director of Product Management, dealing with numerous aspects of product life cycle including (1) idea conception, R&D and prototyping (2) product management (3) designing platform architecture (4) internationalization (5) productionizing code (6) testing (7) security and customer care (8) operations engineering (9) abuse team management (10) legal & community relations. Worked with teams operating 22 markets across 5 continents, travelled globally.

    Researched and developed several prototypes for abuse detection on a site with over 1.5 billion page views and 200 million unique users per month (Yahoo Answers). Implementations have resulted in cost savings to the tune of 3 million dollars per year. This included a community moderation system, which dynamically assigns user reputation and predict human abuse by looking at user interactions. The system achieved a very high abuse judgment accuracy of 98% (10% higher than average customer care accuracy), establishing Yahoo Answers as a role model for moderating large scale User Generated Content. Provided abuse consultancy to Flickr, Groups, Buzz, Messenger & Yahoo Research.

    Individually developed several large-scale data mining and machine learning prototypes using 1000+ node Hadoop cluster. Projects dealt work from analytics, trend detection, pattern mining, abuse/spam detection to search engine components. Acted as the go-to guy for getting data and testing hypothesis for new features. Notable projects include an Automated Question Answerer, which replies fact based answers to English questions by mining web data.

    Recognized for performance with 3 company-wide performance accolades/awards.

left Publications right

  • Koul, Anirudh and Khurana, Udayan, 'Determination of Sequences in Prime Numbers', National Conference on Bioinformatics Computing (NCBC), India, February 2005, pp. 175-177, Best Paper Award.
  • Khurana, Udayan and Koul, Anirudh. ‘Using Patterns to Generate Prime Numbers’. In 3rd International Conference on Advances in Pattern Recognition (ICAPR), United Kingdom, August 2005, pp. 325-334.

left Research Experience in University right

  • Automatic Network Traffic Classification for Peer-to-Peer Traffic [1]
    The research developed methods to classify Peer-to-Peer traffic accurately, with the methods being adaptable to changes in port, payload patterns or protocol. Features from the packet headers were identified, and were used to train machine learning algorithms for classifying network traffic. Generic and application specific rules were also identified.
  • Non Redundant Data Compression Technique [2]
    This new technique does not depend on repetition and produces enhanced results consistently as opposed to existing compression techniques that utilize algorithms requiring repetitive units within data for compressibility. The technique also enables a feature called "Progressive Page Rendering" that allows a file to be opened and viewed even while the file is being downloaded.
  • Determination of Sequences in Prime Numbers [2]
    Generated a Periodic Sequence whose numbers have a very high probability of being prime numbers. This sequence has been utilized as a new deterministic algorithm for faster generation of prime numbers of large magnitudes. The algorithm shortens the time and lowers the time complexity required to compute large prime numbers.
  • Optimized algorithm for generating Magic Squares [2]
    Developed an optimized algorithm to generate Magic Squares and Magic Cubes that lowers the memory and computational resources required along with low time complexity. The algorithm generated Magic Square of the order of 10001x10001, and thus broke the previous record of the order of 3001x3001.
  1 - Supervisor: Dr. Nur Zincir Heywood, Associate Professor, Dalhousie University, Canada
2 - Along with Udayan Khurana, TIET, India

left Academic right
  • M.S. Very Large Information Systems (VLIS), Carnegie Mellon University (2010-2012)
    * GPA: A : 4.0

    * Specializing in Scalability, Machine Learning, Natural Language Processing & Information Retrieval

    * Developed 2 automated Question Answering systems, including one under the guidance of Dr Eric Nyberg (core collaborator to IBM Watson), the other winning 'Most Innovative Question Answering system' from Google for NLP Class Demo. Projects developed include (1) Predicting election outcome by mining Tweets. (2) Wikipedia Infobox Generator Using Cross Lingual Unstructured Text (3) Predicting financial trends using news (4) unsupervised real time spam detection of tweets (5) Personalized audio news with recommended ads (6) Fast & scalable Google Suggest like Ajax backend for large data (run on Google n-Grams)

    * Thorough understanding of Hadoop ecosystem, Cloud computing, Distributed Databases (NoSQL) and processing Big data

  • B.S. Computer Science (with distinction), Dalhousie University, Canada. (2007)
    GPA: A- : 3.75/4 (87%) (2007 Computer Science Sexton Scholar Award)

  • Transferred to Dalhousie University after completing 2 years B.E. Computer Science,
    from Thapar Institute of Engineering & Technology (TIET), Patiala, India (June 2005).

  • Class-XII (1st Division) from ‘CBSE, Delhi, India’ in the year 2003 .

  • Class-X (1st Division) from ‘CBSE, Delhi, India’ in the year 2001.
left Other Accomplishments right
  • Photography
    I am quiet passionate about photography and have a growing online portfolio of heavily viewed Creative Commons licensed photographs on Flickr. Some notables publications where my photos have appeared include the New York Times, Yahoo Homepages, Fox News, ABC News, Lifehacker, Consumerist, The Economist, ReadWriteWeb, Consumerist, Ars Technica, LockerGnome, Air America, Venturebeat, Digital Photography School and MSN Japan Homepage. In total, over 2200+ different websites are using my photographs. My personal photo profile has over 1.2 million page views.. A sample of my work is visible here.

  • At Thapar Institute of Engineering and Technology (TIET), India
    • Intel Ambassador (2004-05)
    • Founder of computing society 'Inquisitive Organization for Technical Aspirants (IOTA).
    • Procter of Hostel B, TIET (2004-05).
  • Head Boy, St. Andrew Scots Public School, Delhi, India, 2002-2003.
  • 15th rank in the 2nd National Cyber Olympiad, India, 2002.
  • Proficient in keyboard / piano, participated in various musical events.
  • Awarded ‘Rajya Purskar' for Scouting from President of India, 1999.