Green leaf

Harsh S - Machine learning developer
Member since:Mar 14, 2020
Profile last updated:Jul 25, 2020
Last activity:Jan. 14, 2021, 7:46 a.m. UTC

Harsh S

ML Engineer

A Machine Learning engineer with experience in Deep Learning and Data Science. Application of analytical skills for problem solving keeps me passionately involved. I am interested in applying technology for real applications and always desire working in a fast paced, test-driven and collaborative engineering environment.

Skills:   Machine learning  ·  Natural language processing  ·  Deep learning  ·  Computer vision  ·  Data Science  · 
Weekly Availability: 12 hours
Mon 18 - 21 10 - 13
Tues 18 - 21 10 - 13
Fri 18 - 21 10 - 13
Sat 18 - 21 10 - 13

Partime Freelancer
Experience: 4+ yrs
Machine learning: 4+ yrs
Natural language processing: 4+ yrs
Deep learning: 3+ yrs
Computer vision: 3+ yrs
Data Science: 3 yrs
Engineer's Devices:

LD Talent Work History

Average response time:
10.6 hours
Avg count of messages / Day:
# Hires / # Interviews :
1 / 1


American Express     Personalisation of offer campaigns     Data Scientist     Employment
Jul 2018 - Present

  • Machine Learning: Built Machine learning models for personalised offers on American Express Cards
  • Data Science: Perform Quantitative Analysis for designing offer construct

Skills used: Data Science, Machine learning     Data Science & NLP     Data Scientist     Internship
May 2017 - Present

  • Machine Learning: Built model for text documents classification using Neural Networks and NLP.
  • Natural Language processing: Developed algorithm for hierarchical clustering

Skills used: Machine learning, Natural language processing

TowardsDataScience | Medium     Blog writing     Contributing Writer     Contract
Mar 2019 - Present

  • Data Science: Authoring articles revolving around Deep Learning, Machine Learning and Data Science in general
  • Machine Learning: Been read by more than 2.5k readers, articles attract 1k+ views monthly with best read ratio of 63%

Skills used: Data Science, Machine learning

Link to the project:
IIT Delhi     Dominant Speaker Recognition in videos     Student     Course Project
Feb 2018 - Present

  • Extracted and prepared frame-wise data of 6 speakers (1500 images per speaker) via youtube using Opencv
  • Computer vision: Used OpenCV's Classi er and a heuristic approach based on consecutive images' correlation to remove noise
  • Deep Learning: Implemented Grad-CAM and t-distributed stochastic neighbor embedding (t-SNE) to visualize deep net

Skills used: Deep learning, Computer vision

Link to the github:
IIT Delhi     Hard Coding Core Work-horses of Deep Neural network     Student     Course Project
Jan 2018 - Present

  • Machine Learning: Hard coded a feed forward neural network from scratch without using machine learning libraries
  • Added Lasso, Ridge regularization, batch normalization and dropout to improve the working of DNN

Skills used: Machine learning

Link to the github:
IIT Delhi     Thesis: Stock prediction using Deep Learning     Student     Course Project
Jan 2017 - Apr 2018

  • Prepared our own novel nancial news data via scraping web pages for several stocks and indexes
  • Natural Language Processing: Experimented with several novel techniques to quantify text, namely, word2vec, tf-idf and sentiment scoring
  • Machine Learning: Combined stock and text data using t-distributed stochastic embedding (t-SNE), fed into LSTM architecture
  • Implemented CNN-LSTM taking several days' news as input to preserve both spatial and temporal component

Skills used: Natural language processing, Machine learning


Indian Institute of Technology Delhi (IIT Delhi)    B.Tech 
Jul 2014 - May 2018


Advanced Machine Learning    IIT Delhi
Jul 2017 - Nov 2017
Skills learnt: Deep learning, Machine learning
Deep Learning    IIT Delhi
Jan 2018 - Apr 2018
Skills learnt: Deep learning, Computer vision, Machine learning