Experience
Resume
Matthew Byrd
Software Engineer @ Epic
In 2021-2022 I was a graduate student at UNC-Chapel Hill working on a M.S. in Computer Science under Dr. Shashank Srivastava. My primary work was concerned with predicting the difficulty of a question based on the text of the question. This work was published and presented at ACL 2022. Alongside my research, I worked with Dr. Tyson Hedrick to develop research software for Biology. During this time, I created a web app supporting complex video analysis that was published at the Society for Integrative and Comparative Biology 2021.
During my graduate studies I also contracted as a web developer for both front-end and back-end work with Caktus Group. I worked with many clients such as Rue Gilt Groupe and Magnit. I also briefly lead a contract with AppOmni. During this time, I was responsible for ensuring our team finished deliverables on an agreed upon time-table, as well as communicating needs between the team and client.
I am currently working at Epic Systems with the Cosmos team with interest in Education Technology and Natural Language Processing
Reach me at: matthew_a_byrd at outlook dot com
Publications
Association for Computational Linguistics (2022)
Society for Integrative and Comparative Biology (2021)
Experience
Web Developer, Caktus Group
January 2022
Highlights:
- Lead a team on a contract to develop a Django backend app that queried user security settings and alerted end-users for incorrect configurations. Added nine new supported apps to our client’s platform.
- Upgraded client Django website from 1.8 to 1.11. Used Docker, AWS for development environments. Utilized GitHub continuous integration tools to deliver a well-tested upgrade that was deployed Fall 2022.
- Primary back-end engineer on a Django/React website that queried private APIs with Django Rest Framework to present potential candidates for a given job position
Description:
Over the years I've worked on many web projects, including the contracts I've taken on with Caktus. You can view some of them here.
Researcher, University of North Carolina at Chapel Hill
May 2020
Highlights:
- First author on paper published at top NLP conference ACL studying the predictability of difficult questions in question-answering tasks. Showed that the difficulty of questions can be predicted using a variety of simple and interpretable natural language features. Selected for oral presentation and presented at ACL 2022.
- First author on in-review paper proposing a method to predict medical error severity. Used Pytorch, Huggingface, Sklearn to develop and interpret complex deep-learning BERT-based classifiers as well as simple bag-of-word models. Achieved .7 F1 score in some tasks.
- Researcher on on-going project on how to use natural language feedback to improve black-box models
Description:
During my years as a researcher, I published a paper and implemented various practical models. You can view my work here
Research Programmer, University of North Carolina at Chapel Hill
August 2019 - December 2021
Highlights:
- Sole author of a Django website with a JS front-end that provided a feature-rich video analysis tool that is now deployed as a Heroku web app and utilized by UNC-CH Biology labs.
- Integrated pre-trained computer vision models to assist in automated video analysis using tensorflow.js.
- Published technical contribution to biology community.
Description: