# Yaser Qazi > Engineering @ Meta Location: San Francisco Bay Area, United States Profile: https://flows.cv/yaser ## Work Experience ### Software Engineer @ Meta Jan 2022 – Present | San Francisco Bay Area ### Software Engineer II @ Alarm.com Jan 2021 – Jan 2022 ### Software Engineer @ Alarm.com Jan 2019 – Jan 2021 ### Undergraduate Researcher @ University of Virginia Jan 2018 – Jan 2019 | Charlottesville, Virginia Area Machine reading research has reached the point where given a text and a reading comprehension question regarding that text, a trained model can specify the starting and ending point in the text that contains the answer to the question. However, there exists the problem of 'unanswerable questions', or questions that are designed to look like they contain an answer in a given text, but in reality cannot be answered by the text. Current models lack in this ability to detect whether or not a given question is actually answerable; this problem forms the basis of my research. - exploring the field of natural language processing by utilizing an application of conversation modeling to reading comprehension tasks - building models that can perform reading comprehension tasks by answering questions relevant to a context paragraph or document - training models to detect adversarial questions whose answers do not reside in given context paragraphs or documents ### Peer Tutor @ Knack Technologies, Inc. Jan 2017 – Jan 2019 ### Software Engineering Intern @ Federal Reserve Bank of New York Jan 2018 – Jan 2018 | Greater New York City Area INTERNSHIP - employed machine-learning and natural language processing algorithms using Python-based libraries to develop a micro-service to effectively test web applications central to the Bank's operations - hosted this micro-service as a Flask-based REST API - documented API using Swagger UI Tools - aided in the creation of a Selenium script which called this API to test web forms FEDERAL RESERVE BANK CRUNCH-A-THON - participated in the annual bank-wide Data Science Crunch-a-thon, used Python to identify trending words and phrases in bank speeches over time by pre-processing all bank speeches of the last three years as raw text and then, assuming a Poisson distribution, identifying those words that crossed a certain threshold, used Tableau to construct time series of these words ### Undergraduate Teaching Assistant @ University of Virginia Jan 2016 – Jan 2016 | Charlottesville, Virginia Area I served as an Undergraduate Teaching Assistant in the Department of Computer Science for CS 2102 - Discrete Mathematics. I held office hours for 3 hours each week for students to come in and ask questions about homework, lectures, or just class material in general. Along with holding office hours, I collaborated with other teaching assistants and the professor to improve the course where needed, and also helped grade student exams. ### Research Intern @ Virginia State University Jan 2016 – Jan 2016 | Virginia State University I led a team of three students to research an application of the change-point detection problem. Change-point detection is concerned with determining whether or not a change has occurred in a system and identifying the time of any such change. It arises in applications such as network security, airport security, meteorology, and many other disciplines. We investigated the problem of detecting the introduction of a point radiation source into a low-radiation environment. Radiation level can be modeled by a Poisson probability distribution, whose parameter changes once a point radiation source is detected in the system. Therefore, after a radiation source has been detected, the probability distribution shifts according to this new parameter. Our goal in the research project was to determine the point in time, with minimal detection delay, when this probability distribution shifted. We employed a variant of the logarithmic-likelihood function to clean our probability distribution data, and attempted to determine the point in time when this likelihood function reached its maximum. To achieve this goal, we developed and tested extensive MatLab scripts by integrating our written code with methods of various statistical libraries. At the conclusion of our research, we presented our findings at an undergraduate conference held at the College of William and Mary on June 28, 2016. ## Education ### Bachelor’s Degree in Computer Science and Statistics University of Virginia Jan 2015 – Jan 2019 ### Master of Science - MS in Computer Science Georgia Institute of Technology Jan 2021 – Jan 2024 ## Contact & Social - LinkedIn: https://linkedin.com/in/yaserqazi --- Source: https://flows.cv/yaser JSON Resume: https://flows.cv/yaser/resume.json Last updated: 2026-03-23