Karan Tyagi

  Education

Northeastern University, Boston, MA			       September 2017 - December 2019
Khoury College of Computer Sciences	               
Masters of Science in Computer Science
GPA: 3.81 / 4.0

Institute of Engineering and Technology, Lucknow, India	                July 2010 - June 2014
B.Tech in Computer Science and Engineering
Distinction with Honors (1st Division)
GPA: 3.68 / 4.0

  Course Work

Large Scale Parallel Data Processing (Hadoop MapReduce, Spark)

Information Retrieval
Search Engine Design - Web crawling, Text acquisition and pre-processing, Indexing and Storage, Link Analysis through PageRank algorithm, User Interaction, Retrieval Models (tf-IDF, BM25, Query Likelihood), Lucene Search, Search Result Evaluation and Ranking Techniques

Data Mining Techniques
Classification, Association rules, Regression, Clustering, Decision Trees, Ensemble methods, Naive Bayes, SVM, Neural Networks, Recommendation Systems

Web Development
Client Side Development : Bootstrap, jQuery, Angular, React, React Native
Server side Development : Spring, Express, Node

Algorithms
Divide and Conquer, Greedy Algorithms, Dynamic Programming, Sorting and Searching, Amortized Analysis, Graphs.

Programing Design Paradigms
Object Oriented Design - Design principles (Java), Functional programming design (Racket)

Database : MySQL, MongoDB

Database Management Systems
Database Design, SQL, RDBMS, JDBC, JPA, Web Services, Big Data

Managing Software Development
SDLC, Functional and structural testing, TDD, Design Principles, Java design Patterns

  Projects

MovieBuddies - Web App for organizing movie parties			  Oct 2018 - Dec 2018
A web app for creating and organizing casual movie parties for friends.
Technologies used: AngularJS | ExpressJS | NodeJS | Mongoose | MongoDB
 https://github.com/karantyagi/Movie-Buddies
 Demo: https://movie-buddies-client.herokuapp.com/home

Chat Server			          			         Sept 2018 - Nov 2018
Implemented a broadcasting chat server akin to Slack with User & Group login, customizable
  profanity filter, real-time wiretapping, message querying, and a command-line interface.
Designed the message persistence feature, and implemented the backend database services.
Technologies used: Java | Spring | Jira | Jenkins | jUnit | Test Driven Development (TDD)

I-Recall : A memory management and recall strengthening app			     Oct 2018
Proposed the idea and system architecture, and built a prototype under 36 hours at
  HackHarvard 2018 hackathon.
Supervised the creation of a prototype and demoed it during the judging rounds
Modeled and analyzed Ebbinghaus Forgetting Curve for our use case.
Created visualizations for topic-wise recall curves and for cumulative recall curve.
Wrote services for updating the visualizations in real time and for initializing, and
  resetting sample data for the demo.
Technologies used: Charts.js | Javascript | Microservices
 https://github.com/karantyagi/iRecall-HackHarvard18

Job Search Made Easy			          			 May 2018 - July 2018
A one-stop solution for searching, posting and applying to jobs
Designed a job portal and implemented it as a RESTful web app for aggregating jobs from 
  APIs like Github Jobs API and jobs posted by registered recruiters
Technologies used: AngularJS | ExpressJS | NodeJS | Mongoose | MongoDB
 https://github.com/karantyagi/job-portal
 Demo: https://job-search-made-easy.herokuapp.com/home

White Board - Course Management System			               April 2018 - June 2018
Online course management system for creating, managing courses and registering students
Admin and Faculty modules were built using jQuery, React using a Spring Boot Java server
  connected to MySQL, deployed on Heroku
Student module was built using the MEAN stack (MongoDB, Express, Angular, Node.js)
Technologies used:  Bootstrap | jQuery | JavaScript | Angular | React | React Native |
  Spring Boot | MySQL | MongoDB | Node
 https://github.com/karantyagi/hackathons-near-me
 Demo: https://kt-whiteboard-angular-client.herokuapp.com/

Information Retrieval Systems			          	      March 2018 - April 2018
Developed a Search Engine in Java with the following features: Query Expansion,
  Pseudo-Relevance Feedback, Snippet Generation, Synthetic spelling-error generation
Implemented query expansion using pseudo-relevance feedback
Dataset: CACM test collection
Created an inverted index after processing the content of the CACM corpus
Four baseline models for retrieval: BM25, TF-IDF, Smoothed Query Likelihood, and Lucene
Evaluated the performance of the different IR systems using
  Precision, Recall, Mean Average Precision, Mean Reciprocal Rank
Technologies used: Java | Apache Lucene | Maven | IntelliJ
 https://github.com/karantyagi/information-retrieval-systems

OvenBot - Display Navigation system and conversational bot for the Blind	   April 2018
Developed a prototype within 18 hours, which allows the blind to easily navigate the
  digital displays of ovens and microwaves, using fingertip detection and beep generation
The frequency and pitch of beeps change as you hover your finger over the target point on
  the digital display
Technologies used: Python | OpenCV | Google Cloud Services - OCR API | Watson Speech API
 https://devpost.com/software/making-a-meal
 Perkins Hacks Making a Meal challenge
 https://github.com/karantyagi/PerkinsHacks18

Yelp Restaurant Recommendations			          	       Feb. 2018 - April 2018
Predicted ratings from reviews using sentiment analysis
Suggested new restaurants to Yelp users, based on predicted ratings of other Yelp users who
  have reviewed similar or the same restaurants as the user
Technologies used:  Python | NLTK | Jupyter Notebook | Sentiment Analysis
 https://github.com/karantyagi/Restaurant-Recommendations-with-Yelp

Emotion Recognizing Music Player			          	        February 2018
A Music system to change songs depending on a person's mood
Detecting users' mood in near real-time, based on their facial expressions
Built under 24 hours at HackBeanPot2018 Hackathon, Boston
Technologies used: Python | Amazon Web Services | Machine Learning | Leap Motion Controller
 https://github.com/karantyagi/emotion-recognizing-music-player

Web Crawler for Wikipedia			          	 	  Jan 2018 - Feb 2018
Crawler for crawling Wikipedia articles using seed URL, keyword and depth parameters with BFS and DFS implementations
Focused crawling for achieving higher topical convergence
Implemented politeness policy by using a delay of at least 1 second between HTTP requests
Technologies used:  Python | Beautiful Soup
 https://github.com/karantyagi/web-crawler

American Sign Language Recognizer			          	        November 2017
An interface for Speech Disabled people to communicate with other people who may not know
  American Sign language, using the Leap Motion Controller
The interface recognizes gestures and converts them into Audio and Text.
Technologies used: Python | OpenCV | Microsoft Azure Emotion API
 https://devpost.com/software/asl-recognition
 https://github.com/karantyagi/American-Sign-Language-interpreter

Squash Game			          				Nov. 2017 - Dec. 2017
Created a ball - racket game where the player plays to maximize his/her time
Developed user interactive animation employing techniques such as Design by Contract, High
  order functions, and recursion based programming
Technologies used:  Racket | Functional Programming
 https://github.com/karantyagi/ball-racket-game

Virtual Mouse			          			      	     June - July 2017
Developed a Windows application using MATLAB which simulated mouse functionality like
  point-and-click and drag-and-drop using live video feed from laptop’s webcam
Compared various color detection algorithms to recognize colored markers on fingertips
Created a module for manual illumination calibration
Extended this project to make a Powerpoint app which allowed switching slides by waving
  your hand, and an image app with pinch zoom gesture functionality
Technologies used:  MATLAB

  Hackathons and Awards

PerkinsHack 2018 | Perkins School for the Blind, Watertown, MA
Won the "Making a meal challenge - Accessibility for the Blind"
Oven Bot : Universal digital screen navigation and conversational assistant for making 
  ovens (and other kitchen appliances) more accessible for the Blind.
Developed a prototype within 18 hours, which allows the blind to easily navigate the
  digital displays of ovens and microwaves, using fingertip detection and beep generation
The frequency and pitch of beeps change as you hover your finger over the target point on
  the digital display
Technologies used: Python | OpenCV | Google Cloud Services - OCR API | Watson Speech API
Know More:  https://devpost.com/software/making-a-meal
 Perkins Hacks Making a Meal challenge
 https://github.com/karantyagi/PerkinsHacks18


HackWITus 2017 | Wentworth Institute of Technology, Boston, MA
Won "Best use of Amazon Web Services" prize
American Sign Language Recognizer: An interface for Speech Disabled people to communicate
  with other people who may not know American Sign language, using the Leap Motion 
  Controller.
The interface recognizes gestures and converts them into Audio and Text.
Technologies used: Python | OpenCV | Microsoft Azure Emotion API
Know more:  https://devpost.com/software/asl-recognition
 https://github.com/karantyagi/American-Sign-Language-interpreter


HackBeanPot 2018 | Boston, MA
Emotion recognizing music player: changes songs depending on a person's mood
Detecting users' mood in near real-time, based on their facial expressions.
Technologies used: Python | Amazon Web Services | Machine Learning | Leap Motion Controller
 https://github.com/karantyagi/emotion-recognizing-music-player


Goju-Ryu Karate									  August 2015
 Green-II Belt (4th Kyu), International Academy of Martial Arts, India
 View Certificate 


Academic Excellence Award							    June 2014
 Silver Medal, 2nd rank in Computer Science and Engineering - 2014 batch
 during under-graduation