Single Page Resume[PDF] | 2 Page Resume[PDF]

Summary

Pursuing Master of Science in Computer Science at Georgia Institute of Technology, Atlanta with a specialization in Machine Learning. Exploring Deep Learning methods in Computer Vision.

Research Interests

Machine Learning, Deep Learning, Computer Vision

Graduate Courses

  • Natural Language
  • Adv. Computer Vision*
  • Artificial Intelligence
  • Computer Vision
  • Computer Graphics

Education

  • Aug 2017 - Present: Master of Science, Computer Science
    Georgia Institute of Technology, Atlanta
    CGPA: 4.0 / 4.0

  • 2009 - 2015: Master of Science(Hons.) Mathematics and Bachelor of Engineering (Hons.) Computer Science
    Birla Institute of Technology And Science, Pilani Goa Campus
    CGPA: 7.85 / 10

Academic Projects

Text Classification, Feb 2018

Implemented lyrics classifiers based on Naive Bayes, Perceptron, Logistic Regression techniques

Sequence Labeling, Mar 2018

Implemented a Part-of-Speech tagger based on Hidden Markov model and BiLSTM - Conditional Random Field models

Dependency Parsing, Apr 2018

Coreference Resolution, Apr 2018

Built a coreference resolution pipeline using Attention-Based-LSTM

Morse Code Recognition, Dec 2017

  • Implemented Morse code recognition system through the use of Hidden Markov Models

Scene Classification with Deep Learning, Nov 2017

Report

  • Built a Convolutional Neural Network to classify scenes into categories
  • Used Transfer learning from the VGG network to classify scenes on the SUN dataset

Surface Reconstruction from Point Cloud Data, Nov 2017

Report | Project Source

  • Water Tight Surface Reconstruction of 3D Point Cloud Data using the Ball Pivoting Algorithm

Automatic Image Segmentation using Expectation Maximization, Nov 2017

  • Implemented automatic image segmentation using K-Means, Gaussian Mixture Models employing the Bayesian Information Criterion

Scene Recognition with Bag of Words, Oct 2017

Report

  • Developed a scene recognition pipeline with Bag of SIFT and linear SVM classifier

Local Feature Matching, Sep 2017

  • Developed a local feature matching algorithm employing Harris Feature Point Detector and implemented the SIFT algorithm for local feature descriptor

Face Detection, Sep 2017

Report

  • Developed a local feature matching algorithm employing Harris Feature Point Detector and implemented the SIFT algorithm for local feature descriptor

Dead-end Isolation Game Player, Aug 2017

  • Developed a dead-end isolation game player based on Minimax Algorithm using Alpha-Beta pruning

HTTP Client using Socket Programming, April 2014

Project source

  • Developed an HTTP Client using Socket Programming in C. This client was designed to GET an HTML page pointed to by the given HTTP URL and all its associated images.
  • Additionally, the application, decoded images encoded using the base64 encoding technique and generated the equivalent PNG images

Analysis of finite sets sampled from Euclidean space quasi-ordered by comparison relations, Jan 2014 - May 2014

Adviser: Dr. Ramprasad Joshi

  • Study of embedding weighted graphs from Rn space to R3 space to help solve and visualize n-ary equations with n-variables

Raytracer, August 2013

Project source

  • Implemented a Raytracer for lambert materials in C++ using OpenGL 2.0 API representing the polygonal mesh using a winged-edge data structure

Subdivision Surface Modeling of Polygonal Meshes, Jan 2013

Adviser: Dr. L. Gudino
Project Source

  • Study of data structures used to describe polygon meshes, polygon subdivision conversion algorithms, and theoretical concepts of surface continuity
  • Developed an application in C using the OpenGL API and implemented the Catmull-Clark and Doo-Sabin subdivision conversion algorithms

Work Experience

Uber Technologies, California

Software Engineer Intern May 2018 - July 2018

CatchME (Catch Map Errors)

Technologies Used: Apache Spark, Java
  • Prototyped Hidden Markov Model-based techniques to find Map errors using drivers’ GPS traces

MapCrunch Reliability

Technologies Used: Apache Spark, Java, Docker, Graphite
  • Built spark-based tools for Map Reliability
  • Improved existing pipeline performance by optimizing Spark usage

BlueJeans Networks, Bangalore

Senior Software Developer, Core Platform Team
May 2015 - Jul 2017

Next Generation Platform

Technologies Used: Java, NodeJS, Spring-Boot, Hystrix, Ribbon, OkHttp, NGINX, Feign, Graphana, Docker, Wowza Streaming Engine, AWS DynamoDB
  • Re-engineered the application design to enable a pure micro-service architecture
  • Refactored the existing monolithic backend codebase to facilitate deployments as microservices, involving migration of the Google Guice based framework to Spring Boot
  • Migrated the existing technology stack to Amazon Web Services Cloud.
  • Using Infrastructure-as-Code, the system can now be re-instantiated in any AWS data center
  • The improved performance, scalability and reliability resulted in an increase in the Monthly Uptime Percentage from .99 to .9999
  • System capacity increased from 5k users to 45k
  • System load limit increased from 500 API requests/sec to 100k requests/secs

Social Media Gateway

Technologies Used: NodeJS, Typescript, coa, Wowza Streaming Engine, AWS Lambda, AWS ECS, AWS DynamoDB_
  • Designed, developed and deployed a AWS Lambda based NodeJS solution to enable live streaming video conferences into RTMP entry points achieving a time to market target of 14 days
  • The solution is now integrated with Facebook Live video and has been tested successfully with YouTube

Autoscaler Service

Technologies Used: Java, Wowza Streaming Engine SDK, AWS EC2
  • Developed a transcoder-instance autoscaling system to enable auto provisioning of AWS EC2 instances based on real-time usage patterns
  • The service equipped the system with the ability of hitless upgrades and enabled blue-green deployments and reduced server usage costs by 55%

BlueJeans Networks, Bangalore

Software Development Intern, Core Platform Team
Jul 2014 - Jul 2015

Technologies Used: Java, Gradle, Docker, Kubernetes
  • Completed a POC for deploying the Primetime backend stack on AWS EC2 using Kubernetes
  • Developed tools for Web UI Automation to enable multi-browser, multi-node, multi-device concurrent testing
  • Designed and developed a tool to simulate high server loads for benchmarking and stress testing

National Aluminium Company Limited, Bhubaneswar

Software Development Intern, Systems
May 2012 - Jul 2012

Technologies Used: HTML, CSS, ASP.NET
  • Redesigned and developed the ASP .NET based internal website

Ojaswi Tech, Pune

Graphic Design Intern
Jun 2011 - Aug 2011

Technologies Used: Adobe Photoshop, Adobe Illustrator, Adobe Dreamweaver, Adobe After Effects, Adobe Flash
  • Developed UI mocks for a client websites and developed UI elements standards.
  • Developed Marketing Presentation Videos for clients

Certifications

  • Machine Learning
    Instructor: Andrew N G
    Authorized by Stanford University and offered through Coursera
    Certificate Link

  • Java
    SEED InfoTech, Pune
    Completed the training with Grade: A+

Teaching Experience

  • 2012 - 2013: Teaching Assistant for Computer Graphics under Dr. Lucy Gudino
  • 2012 - 2013: Teaching Assistant for Human Computer Interaction under Dr. Mangesh Bedekar
  • 2013 - 2014: Teaching Assistant for Computer Graphics under Dr. Lucy Gudino
  • 2013 - 2014: Teaching Assistant for Creative Multimedia under Sreejith V

Professional Awards

  • Quarterly Award
    2016 - 2017 Q2
    BlueJeans Networks