Pranav Shrestha

Education

Columbia University in the City of New York Aug 2016 - May 2020
  • B.Sc. in Computer Science (Intelligent Systems) GPA: 4.15
  • C.P. Davies Scholar, Columbia University Scholar Program. Fu Foundation School of Engineering and Applied Sciences
  • Coursework: Artificial Intelligence, Machine Learning, Statistical Inference, Deep Learning, Operating Systems, Natural Language Processing, Analysis of Algorithms, CS Theory, Advanced Programming (C/C++), Quantum Computing

Research

Qualcomm, Corporate Research and Development May 2019 -
  • Benchmarking recurrent model compression on ARM with open-source optimization tools (Tensorflow Lite and TVM)
Piazza Chatbot, Columbia University Aug 2018 – Apr 2019
  • Developing a generalizable answering bot, designed for massively open online courses (MOOCs)
  • Explored and employed various NLP modelling techniques, centering on language models like Google’s BERT
Secondary Protein Structure Prediction, Columbia University Nov 2018
  • Achieved state of the art ensemble accuracy on protein structure prediction by implementing a bi-directional GRU model with cascading convolutional blocks
  • Compiled open-source repository and co-authored the paper presented at the NeurIPS 2018 workshop - ML for Molecules and Materials (arXiv 1811.07143)

Experience

Qualcomm, Software Intern 2017-2018

Data Team ChromeOS Project

  • Extended Qualcomm’s test framework to ChromeOS by implementing a C++ kernel module for Chromium OS

Data Team Email Reporting

  • Streamlined team productivity and reporting with a full-stack task flow Web UI and email notification system

Drone Computer Vision Team Intern

  • Fully automated a scalable cross-platform continuous testing framework with support for Git and Perforce
  • Supported varying dependencies and branches using Python, C/C++ and ndk-build systems
Liquidaty, Spring Technological Intern Jan 2017 - Apr 2017
  • Completely revamped the web portal for auto-financing analytics focusing on a smoother functional UI and optimized backend processing using Python, Bash Scripting, and Node.js
  • Implemented secure user management and notification systems using Node.js Express and Passport

Achievements and Positions

HackMobile Winner, Qualcomm Jul 2019
  • Developed a low-latency speech translation system using Azure Speech SDK for a gesture-enabled translation app
Columbia Artificial Intelligence (COMS 4701) TA Jan 2019 - Jul 2019
  • Updated course by porting legacy assignments to Python 3 and designing specs/code for a new Keras assignment
  • Improved avg Piazza response time from 60+ to 19 mins and held office hours and recitations for 200 students
BlindData aptitude test Jan 2019
  • Achieved perfect scores on the BlindData adaptive programming test to win the $5000 grand prize (99th percentile)
HackMobile Finalist, Qualcomm Jun 2018
  • Selected as a top 10 finalist by implementing an ultra-low cost, line-of-sight data transfer system using visible light
Google online foo.bar challenge completed
Qualcomm monthly poker tournament winner
Finalist for Microsoft’s Last Slice challenge (top 1000)
CIE A-Levels
  • Top in the world in AS Math (2016)
  • Top in Nepal in A-Level Biology, Chemistry, Physics, Math (2017)