Ekrem Guzelyel

Research Assistant

Machine Learning Lab at Illinois Tech•  October 2018 - April 2020

Research: “Learning with Rationales for Document Classification” • Samsung project for extracting ground truth rationales with Conditional Random Fields (CRF). Research: “Data Categorization for Transparent Text Classification” • Automated labeling and preprocessing of text documents using Machine Learning. • Trained multiple Machine and Deep Learning classifiers to identify reasoning.

Lead Software Developer Intern

Blocklight•  May 2019 - August 2019

BAP: “Business Analytics Tool for Small Businesses” • Managed the development team of 5, and reached to the MVP. • Improved the Django based web-app using Python, and restructured MySQL database. • Integrated third party APIs such as Google Analytics API, Facebook Graph API, django-allauth, and Shopify.

Engineering Practicum Intern

Google•  May 2018 - August 2018

Project 1: “Improving Targeting Quality using TensorFlow Model” • Implemented a negative filtering for Google Search Ads using Deep Neural Networks. • Created automated experiments to tune TensorFlow model performance. • Observed significant improvements of quality (7.2%) over traditional logistic regression model. Project 2: “Model Age Monitor” • Created, tested and launched an online monitoring system that successfully catches old models and executes on-call alert using Python.

Illinois Institute of Technology, Chicago

Artificial Intelligence, MS•  August 2016 - December 2020

Masters in Artificial Intelligence Bachelors in Computer Science Minor in Applied Mathematics