Curriculum Vitae
Education–Research–Projects–Teaching–Standardized Tests–Work–Activities
#####Birla Institute of Technology and Science, Pilani [Aug ‘11 - Jul ‘15]
Bachelor of Engineering (Honors), Computer Science
GPA | 8.54/10.0
######Senior Thesis – Fuzzy Matching for People Search and Neighbourhood De-duplication
Mentors | Advanced Development Group, MSR India & Prof. Yashvardhan Sharma
#####Research Fellow Microsoft Research, India [Aug ‘15 - Present]
Mentors | Monojit Choudhury & Kalika Bali
Computational models for code-mixed text
-- Word-level language identification and code switch-point detection
-- Pragmatic and social functions of code-mixing
-- Analysis of code-mixing geographically and event/topic-wise
#####Research Intern Microsoft Research, India [Jan ‘15 - Jun ‘15]
Bachelor’s Thesis work with the Advanced Development Group
-- People Search – Developed retrieval and ranking techniques for fuzzy matching of misspelled “people” queries in e-mail search.
-- Neighborhood De-duplication – Detecting duplicates among 500,000 unstructured and phonetically variant Indian neighborhood descriptions (without geocodes) was tackled with fuzzy matching, synonym learning and subset grouping.
-- Both projects involved prototyping and experimentation with different matching metrics and heuristics. [Skills – C#, SCOPE]
References – Udupa, Raghavendra, and Shaishav Kumar. “Hashing-based approaches to spelling correction of personal names.” Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics, 2010.
#####Intern, Bing Ads Microsoft IDC [May ‘14 - Jul ‘14]
Mentor | Mehul Parsana
-- POS-tagging search queries with Conditional Random Fields and search logs.
-- The model developed resulted in a 6% increase in accuracy over existing methods.
-- Experimented with location entity tagging in queries using iterative learning.
-- A POS-tag discovery mechanism for unknown words was developed and tested with a range of training lexicon sizes.
References – [1] Bendersky, Michael, W. Bruce Croft, and David A. Smith. “Structural annotation of search queries using pseudo-relevance feedback.” Proceedings of the 19th ACM international conference on Information and knowledge management. ACM, 2010.
[2] Ganchev, Kuzman, et al. “Using search-logs to improve query tagging.” Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics: Short Papers-Volume 2. Association for Computational Linguistics, 2012.
#####Google Summer of Code MIT Media Lab [Jun ‘15 - Aug ‘15]
Google Summer of Code is a global program that offers students stipends to write code for open source projects.
-- Worked on MIT’s App Inventor, an open-source project that is a beginner’s introduction to programming and app creation. [Project fork]
-- Bug fixes that involved hacking large amounts of code across multiple technologies.
-- Worked on SVG-capture and Undo features for the Blocks Editor. [Skills – JavaScript]
#####NewsLine – Event-Centric News Clustering [Oct ‘14 - Dec ‘14]
Clustering of news articles that report various facets of the same event.
-- Built a web crawler to retrieve articles from news sources [Skills – Python, Scrapy]
-- Used Hierarchical Agglomerative Clustering over TF-IDF cosine similarity, maximizing the silhouette score metric to find optimal clustering [Skills – Scikit]
#####Cellphone Microscope – Cheap and Portable Malaria Detection [Jan ‘14 - Dec ‘14]
-- Built a hardware attachment for a cellphone that can magnify images to microscopic extent using three lenses.
-- Developed an image classifier to identify the presence of malarial cells in the magnified images. Experimented with simple thresholding and bag-of-words on pixels. [Skills – MATLAB]
-- Challenges primarily rose from little available training and test data.
#####“Postmodern” – The State of Contemporary Knowledge [Aug ‘14 - Dec ‘14]
Adviser | Prof. Hari Nair
-- Studied Jean-François Lyotard’s A Postmodern Condition as well as works of other Postmodern philosophers.
-- Heavily analysed the changing state of knowledge with increasing computerisation.
-- Investigated Massive Open Online Courses (MOOCs) and other contemporary forms of knowledge transfer.
#####Object Tracking using Segmentation [Jan ‘14 - May ‘14]
Adviser | Scientist Sanjay Singh
-- Built an extension to the basic CAMShift tracker in OpenCV which allowed modifiable threshold values and noise reduction.
-- Implemented a single object tracker based on subtraction of frames and achieved a rate of 100 frames per second.
#####Graph-based Methods for Image Segmentation [Aug ‘13 - Dec ‘13]
Adviser | Prof. Mukesh Kumar Rohil
-- Implemented Prim’s and Kruskal’s spanning tree algorithms for image segmentation in C++.
#####Transformation of Curves [Mar ‘14 - Apr ‘14]
#####Doppler Effect Simulator [Mar ‘14 - Apr ‘14]
#####Compiler for a Matlab-Like Toy Language [Feb ‘14 - Apr ‘14]
#####Page Replacement in PostGreSQL [Mar ‘13 - Apr ‘13]
#####Gandhi’s Care of the Self – Dietetics and Other Thoughts [Sep ‘14 - Dec ‘14]
#####Francis Fukuyama, The Origins of Political Order – A Review [Feb ‘13 - May ‘13]
#####Napoleon’s Impact on Europe [Sep ‘12 - Dec ‘12]
#####Co-Founder & Lecturer getPlaced() [Dec ‘14 - Present]
#####Lecturer Computer Programming Special Interest Group [Aug ‘14 - Dec ‘14]
#####Teaching Assistant Logic in Computer Science [Aug ‘13 - Dec ‘13]
#####Graduate Record Examination (GRE) [July 24th, 2014]
Quantitative Analysis | 169/170 Verbal Reasoning | 170/170
Analytical Writing | 5.0/6.0
#####Test of English as a Foreign Language Internet-Based (TOEFL) [October 24th, 2015]
Total Score | 120/120
#####Lead Developer Lost Child Art Studios [Nov ‘13 - Feb ‘14] Lost Child Art Studios is a startup that offers complete design and web development solutions.
#####UX Intern Tune Patrol Media Labs Pvt. Ltd. [May ‘13 - Jul ‘13]
Tune Patrol is an online marketplace for independent Indian musicians to expand their audience.
-- Developed the user interface for music concerts and song playlist features on the product website.
-- The playlist feature built resulted in a 300% increase in number of song plays.
#####Software Development Intern Tata Consultancy Services [May ‘12 - Jul ‘12]
Tata Consultancy Services is a multinational information technology consulting and solutions firm.
-- Studied Linux operation and SQL extensively. Built a database management system in C++.
#####Web Development Intern InfoToros [May ‘11 - Jul ‘11]
InfoToros is a startup that provides both business and personal web solutions.
Piano, swimming, 1SCUBA, barefoot running.
###Other