Sidharth Malhotra

I am a tech-agnostic, and flexible full-stack software engineer with 4+ years of experience in developing scalable applications from initial stages to the final release. I have a diverse development background working with different technologies on the web, and as well as deploying applications on different cloud platforms. I am a goal-oriented problem solver with a growth mindset.

My interests lie in bridging the gap between software engineering and machine learning by applying ML approaches to solving software development and infrastructure problems, efficiently using and customizing off the shelf ML services to build state of art software applications.

Projects

• Tracking Fraudulent Blockchain Transactions using Apache Spark and HBase
  A proof of concept application that uses Apache Spark to crunch the Ethereum transaction data to find absurd transactions and the trail of those transactions based on the sender’s Ethereum wallet address submitted through a user query. Also, compared the performance and speedup for hash and shuffle join and joins using HBase for finding interesting structures in the given graph problem.


• Solving The Royal Game of Ur using Reinforcement Learning
  Created an AI gym simulator for strategic Game of Ur using the MDP approach and used that simulator to train a multi-player reinforcement learning agent built-in python to learn the gameplay and best strategies for different game situations using distinct RL methods including Q-Learning, Monte Carlo, Expected Sarsa.


• Protein Secondary Structure Prediction using Deep Learning
  Trained deep MLP and RNN built using PyTorch, to predict the location of alpha-helices (secondary structures) in protein sequences obtained from human and mouse proteins. Also, compared the similarity between the protein structures of the two organisms by predicting alpha-helices in human proteins using the model trained on mouse proteins and vice-versa.


• Contriver - Image Recognition based Plugin for Twitter Ad Campaigns
  A Twitter plugin that uses Image Recognition System which collects images from your past tweets and based on the object types in the images, suggests what kind of images the user's audience prefers the most, hence providing marketers the power to create highly engaging content for their customers. Built using Symfony, MySQL, Twitter and Clarifai APIs


• Inspect Mail - Filter for Disposable Emails
  A REST-API service, built using Symfony & Angular, for third party applications to filter disposable or spam email addresses before registering them to their system, thus increasing lead generation through email marketing


• Blockvitae - SPA powered by Ethereum Blockchain
  A decentralized online portfolio application built using Ethereum Smart Contracts and Angular. Blockvitae allows users to create their online portfolio by adding personal details, work experience, social accounts, projects and education

Contact

Liked my work? Or have some feedback for me? Feel free to shoot me an email and I will get back to you within 24 hours.