About Me

Siddharth Arora

A machine learning enginner with a background in software development, skilled in creating robust, scalable data pipelines and utilizing machine learning techniques to derive critical insights.

My Career

Realtor.com | Toronto, Canada

  • Designed, developed and deployed a Machine Learning (ML) pipeline to forecast monthly leads in US zip codes for next year inventory planning decreasing complexity, reducing run time, and improving accuracy by 20%.

  • Performed ad-hoc analysis tasks and communicated findings to key decision makers to help facilitate the decision making process.

  • Performed code and design reviews to ensure a healthy codebase and prevent bugs, all while fostering knowledge sharing and collaboration.

June 2021 - Present
Senior Machine Learning Engineer

Intact Labs | Toronto, Canada

Working in a team of 7

  • Built a Machine Learning(ML) pipeline to predict business retention to maximize pre-determined objective functions like profit, growth in units, growth in premium, etc.

  • Monitored and diagnosed models in production to examine for ML drifts and took measures to ensure pre-determined desired level of performance.

  • Worked closely with Actuarial Analysts and AI Developers in an Agile setting to ensure production quality software.

June 2020 - Present
Data Scientist II

Stradigi AI | Montreal, Canada

Working in a team of 9

  • Create data pre-processing and transformation libraries for data cleaning and manipulation using python for Stradigi's AI platform.

  • Benchmark and analyze Stradigi’s Machine Learning algorithms against the current state of the art to ensure cutting edge technology.

  • Build Machine Learning (ML) pipelines for Straidgi's AI platform, from data ingest through to solutions for specific use cases like sentiment analysis, document classification, name-entity recognition, image classification, image segmentation, etc.

  • Work closely with Researchers and AI Developers in an Agile setting to ensure production quality software, and efficacy of Machine Learning models being used.

Nov 2018 - May 2020
Data Scientist

Unbounce | Vancouver, Canada

Working in a team of 4

  • Applied survival analysis methods and techniques to determine key factors in customer churn.

  • Automated the data pipelining infrastructure to extract and transform data into usable and integrous components.

  • Utilized easily decipherable visualizations to report findings to people with little statistical knowledge or non-technical background.

April 2018 - Jun 2018
Data Science Intern (Industry Capstone Project)

Masters of Data Science | University of British Columbia, Canada

A 10-month intensive program focused on developing statistics, computing, and machine learning expertise.

Sep 2017 - Jun 2018
Student

IMS Health (now IQVIA) | Auckland, New Zealand

Working in a team as well as individually

  • Developed and updated functionalities for IMS Health’s Nexxus Mobile Intelligence - a CRM solution, that helps drive customer engagement for live sciences industry using C#, JavaScript, XML.

  • Performed maintenance and bug fixing on the existing code base to ensure production quality software.

  • Participated in daily stand-ups and bi-weekly sprint meetings in an Agile setup and developed interpersonal skills.

Sep 2013 - Feb 2016
Software Engineer

Orion Health | Auckland, New Zealand

Working in a team of 2

  • Developed and integrated new custom defined Orion widgets as a proof of concept for Orion Health products using YUI - a JavaScript and CSS library.

  • Designed and developed test cases for functionalities of Orion Health's Web software to ensure verification and validation, and make testing similar components easier for other testers.

  • Automated the test cases using Sahi - a web automation testing tool.

Nov 2011 - Feb 2012 and Nov 2012 - Feb 2013
Intern

Bachelor of Engineering (Software) | University of Waikato, New Zealand

Aug 2009 - Aug 2013
Student

My Skills

My Projects

Signite - Learning sign language through Microsoft Kinect

Allows instructors to teach signs to the software, which in turn can evaluate students on their performance of the same sign. Tools and technologies mainly used were C# for development, and Hidden Markov Models - a statistical model, for modelling each sign. Visit repo for more details. We also got mentioned here.

Twitter Sentiment Analysis - Analyze and classify tweets as either positive or negative

Queries tweets through the Twitter API using Python and applies classification algorithms to categorize tweet as either positive or negative sentiment. Visit repo for more details.

Wine Selector

An R shiny application developed to compare wine quality and price per country. Here is the link to the app. Visit repo for more details.

TopGearElectrified

A group software development project, developed for an undergraduate course. A multiplayer on-rails 3D car Racing simulation, developed in C#, using the XNA framework. Visit repo for more details.

Social Standards and Salary Inclinations Study

A study conducted as a part of Masters course to determine whether a person's social standards are correlated with a person's expected salary. Collected responses from over 100 respondents through SurveyMonkey. Utilised R for data preprocessing and wrangling. Used logistic regression for modelling the data. Visit repo for more details.