Piyush Papreja




I am a recent Compute Science graduate student from Arizona State University. Currently, I am exploring the field of Music Information Retrieval, focused on playlist representation and recommendation. This blog is my journal as I explore the mystical world of binary and try to create something meaningful.

When not wrangling with the 1s and 0s, I’m likely reading blogs, running or playing guitar.


  • Artificial Intelligence
  • Music Information Retrieval
  • Recommendation Systems
  • Web


  • MS in Computer Science, 2019

    Arizona State University

  • BTech in Information Technology, 2013

    University School of Information and Communication Technology





Jan 2020 – Present Dallas, TX, USA


Center for Coginitive Ubiquitous Computing, ASU

Jun 2019 – Present Tempe, AZ, USA

Research Assistant

Center for Coginitive Ubiquitous Computing, ASU

Jun 2018 – Jun 2019 Tempe, AZ, USA

  • Devised a technique to optimally represent music playlists for music recommendation, using deep learning.
  • Partnered with a fitness-based company for music recommendation for workouts.
  • Built a website for collecting speech data from people with Spasmodic Dysphonia to create a dataset for training the speech-based AI algorithms.

Software Developer

ASU Decision Theater

Jun 2018 – Aug 2018 Tempe, AZ, USA

  • Prototyped a light-weight socket server framework to showcase the capabilities of Chainbuilder, Decision Theater’s core product serving prospective policymaking clients, using Python Django.
  • Designed reusable UI components for the Chainbuilder framework using HTML5, jquery, and D3.js

Software Developer

CWPS Lochbridge India Pvt Ltd

Aug 2015 – Jun 2016 Gurugram, India
Built the backend system for smart-car infortainment systems.

Software Developer

Newgen Software Technologies Ltd

Jul 2013 – Jul 2015 Noida, India
Developed Business Process Management based applications for two national banks in Africa, automating their workflows for processes like account opening, using Java.

Recent Posts

Building Music Playlists Recommendation System

Content taken from our paper titled “Representation, Exploration, and Recommendation Of Music Playlists”

Way Finder

Building an indoor mapping solution using Nodejs and A-star algorithm.

Extending OMPL Plugin for VREP

Making modifications to the motion-planning algorithms implemented in V-REP and testing custom versions.


Analyze my music

Music analysis using Spotify API.


Playlist Recommendation Engine built using sequence-to-sequence learning.

Recent Publications

Quickly discover relevant content by filtering publications.

Representation, Exploration, and Recommendation of Music Playlists

With an aim towards playlist discovery and recommendation, we leverage sequence-to-sequence modeling to learn a fixed-length …