I am currently working at Mercari, Inc. as Machine Learning Tech Lead in Tokyo, Japan. My areas of interest include creating production grade distributed machine learning systems as well as research in Deep Learning, Natural Language Processing, Computer Vision & Speech Recognition. I am an alumni of Indian Institute of Technology, Delhi (batch of 2019) with bachelor’s degree in Electrical Engineering. Feel free to contact me on twitter to have a chat.

I co-host a bi-weekly podcast, Tech Culture Podcast with Kaustubh where we talk about emerging startups, interesting tech products and business trends. Follow our podcast twitter account to stay updated.

Click here to see my resume in pdf format.

Areas of Interest


Programming Languages
Python, Golang

ML Libraries & Frameworks
Tensorflow, Keras, PyTorch, NLTK, Scikit-Learn, Open-CV, Pandas, Numpy

Tools & Platforms
Kubernetes, CircleCI, Docker, Datadog, Google Cloud Platform, gRPC, Git, Sentry, BigQuery, MySQL, MongoDB


Indian Institute of Technology (IIT), Delhi
July 2015 - May 2019 // New Delhi, India
Bachelor of Technology in Electrical Engineering


  1. Data Driven Sensing for Action Recognition using Deep Convolutional Neural Networks
    Lecture Notes in Computer Science, vol 11941. Springer, Cham // Dec 2019
    • Developed a novel data-driven under-sampling method using sub-pixel convolutional layers and integrated it with Inflated 3D ConvNet for action recognition
    • Successfully performed action recognition on both UCF-101 and HMDB-51 datasets at multiple (including very high) under-sampling ratios with small drop in accuracy
  2. Compressive Sensing Based Privacy for Fall Detection
    Lecture Notes in Computer Science, Springer // Dec 2019
    • Developed a privacy preserving fall detection framework based on block based compressive sensing and deep learning which works with wide variety of sensing matrices
  3. Artificial Neural Network based Controller Design for SMPS
    IEEE Xplore Digital Library // Oct 2019
    • Designed a controller using neural network for half-bridge converter based SMPS to replace conventional PID controllers
  4. Few Shot Speaker Recognition using Deep Neural Networks
    Preprint // Apr 2019
    • Developed a few shot speaker identification framework using deep convolutional neural networks with prototypical loss
    • Performed speaker identification and few shot speaker identification tasks on Voxceleb dataset using Capsule Network, VGG and ResNet34 architectures
    • Showed generalization capability of the networks on both tasks by performing experiments on VCTK Corpus


English, Hindi, Japanese


Table Tennis, Listening to podcasts, Kayaking, Hiking, Video Games

List of my favorite podcasts

rss facebook twitter github gitlab youtube mail spotify lastfm instagram linkedin google google-plus pinterest medium vimeo stackoverflow reddit quora quora googlescholar