Akanksha Tyagi


Software Engineer

Get in touch: akankshatyagi18@gmail.com

About


Hi there! Welcome to my personal website. I currently live in LA and am pursuing a Master's degree in Computer Science at USC. The journey has been great so far and I will be graduating in December 2018. I am also working as a Software Engineer Intern at GoLinks and am responsible for managing backend of the application. While continuing to explore the field of CS, I have worked on projects based on Web Technologies, Deep Learning, Operating System.

Portfolio


Experience


Software Engineer Intern
GoLinks, California
June 2018 - Present
  • Responsible for maintaining the backend of application
  • Designed REST API to provide golinks statistics to users
  • Implemented a module to create browse cookie to uniquely identify a login activity
Decision Scientist
Mu Sigma, India
Sep 2014 - Sep 2016
  • Helped Fortune 500 clients to solve their business problems by providing analytical solutions
  • Developed an end-to-end module using Python for providing social media analytics
  • Designed user satisfaction framework to help IT of the largest insurance company in US

Education


Master's in Computer Science

University of Southern California
Jan 2018 - Present

Bachelor's in Computer Science

University of Pune
Aug 2010 - May 2014

DR Web Portal


  • Designed and developed a web portal for Directed Research Course (CSCI 590) at USC to allow clients, coordinators, students communicate with each other.
  • It allows students to look at list of projects offered by clients and coordinators and apply to the ones that interests them.
  • Students can edit their profile to showcase their skills and upload resume.
  • Clients / Coordinators can review students' profiles and send out offer letters to them.
  • Coordinators can add / delete students from class.
  • PHP
    MySQL
    HTML
    CSS
    JQuery
Code on Github Close Project

Travel and Entertainment


  • Built a web application that enables users to search for nearby places around their current location or any other specified location
  • Provided feature to bookmark any result and view it at later time
  • Added feature to view directions walking / driving / cycling directions to the location
  • Used Google API and Yelp API to provide photos and reviews of the place
  • Hosted the application on Google Cloud
  • NodeJS
    HTML
    Bootstrap
    CSS
    AngularJS
Code on Github Close Project

Multi View Image Generation


  • Generated synthetic 2D images of an object in a target viewpoint from single input image using deep learning models
  • Built three deep learning models
    • Autoencoder
    • Autoencoder with Pose information
    • Autoencoder with adversarial training
  • Tested above models on ShapeNet Dataset (mug, car, chair)
  • Python
    TensorFlow
    Linux
Code on Github Close Project

Image Classification


  • Designed a Convolutional Neural Network model to classify images present in CIFAR-10 dataset
  • Experimented with various machine learning techniques to improve accuracy of the model
  • Achieved an accuracy of 70% with the help of Batch Normalization and SoftMax Cross Entropy Loss
  • Python
    TensorFlow
Code on Github Close Project

Weenix OS


  • Implemented processes, threads along with synchronization primitives for Weenix in C
  • Implemented virtual file system and virtual memory system
  • C
    GDB
Close Project