Pushkar Sabharwal

Experience

• 

  University of Alberta

  Apr 2020 - Dec 2020

  Undergraduate Research Associate

  • Conducted literature reviews of over 15 machine learning in Education research papers• Underline the limitations encountered by the researchers in their experimental method and come up with solutions• Reproduce the results produced by the paper by implementing algorithms and experimental method in the research paper using TensorFlow Keras• Co-authored a research paper - Lu, C., Ganguly, P., Sabharwal, P., Karimi Abdolmaleki, M., Southcott, J., Jin, H.-Y., González Esparza, L., & Cutumisu, M. (2021). A review of automated-feedback generation in Intelligent Tutoring Systems. Paper. 49th of the Canadian Society for the Study of Education (CSSE) Conference. Edmonton, AB, May 29 - June 3, 2021. Show less

• 

  Intuit

  Jan 2021 - Aug 2021

  Software Engineer

  • Designed and implemented API for Adminportal to obtain REST access to search pinning and fragments data in AWS DynamoDB tables• Developed a Dynamic Site crawler using Selenium, headless chrome and deployed the crawler to an AWS Lambda function to scrape over 50 different dynamic web pages• Built Search query autocorrect model using Tensorflow and Keras to check for mistakes in the query• Worked with AWS, Jenkins, splunk to setup environments with automated unit, integration testing and deployment pipelines Show less

• 

  Intuit

  May 2022 - Dec 2022

  Software Engineer

  • Orchestrated Search Widget playground used intuit-wide to tweak configurations of the search widget in a GUI playground interface• Leveraged Slack EventsAPI to design and develop a slackbot to cater to scheduling requests on slack using AWS Eventbridge• Created and maintained pipelines to introduce the feature flag functionality to progressively expose Search Platform’s new features• Researched and fixed security issues within organizations’ repositories introduced via third party libraries and improved test coverage to over 95% Show less

• 

  Health Gauge

  Nov 2023 - Dec 2023

  Software Engineer
• 

  WCB-Alberta

  Jan 2024 - now

  Data Analyst
• 

  MedWatch Technologies, Inc.

  Mar 2024 - now

  Machine Learning Engineer

  • Designed, trained, and implemented LSTM, CNN based Neural Networks to ascertain blood glucose measurements from over 10,000 Near-Infrared Reflectance Spectroscopy data readings with over 90% accuracy.• Conducted feature analysis, implemented autoencoders to discern intricate skin topologies using signatures, subsequently integrated findings into blood glucose prediction models, resulting in enhanced predictive performance.• Employed SMOGN Oversampling to balance the dataset by generating synthetic data points, enhancing the representation of minority classes and improving model generalization.• Contributed to the end-to-end full-stack development and debugging of a comprehensive Web application, enabling signal extraction from patches Show less