Developed a cyber-security text classification web system. Afterwards, I maintained this version of the system (R, Shiny, PostgreSQL + word2vec NLP) while designing the new version (Python + LLMs). Additionally, I'm part of the research team that designs and studies the theoretical constructs underpinning this system.

Developed a B2C web app (Vue.js 3 + Vuex + Firebase)

Project 1 (2019-2020): Applied econometric methods to investigate peer-economy and users' behavior on the Stack Overflow platform. More specifically, I used a Hidden Markov model to capture and investigate SO users' lifecycle stages, in order to examine how transitions between these stages affect user activity.Project 2 (2020-2021): Designed and developed Fusion, a web system for improved collaboration between domain experts and AI for tasks of text classification in the areas of cyber-security. By leveraging human-machine reciprocal learning, the system offers improved performance and explainability of the generated ML classification models. Tested the system on two case studies. First to identify illegal drug activities on Darknet forums, and second to identify expert hackers on a cybercriminal Darknet forum.

Conducted research focusing on the interplay between developers, tools, their activities, and how it affects collaboration and communication. My research formed a theory of distributed knowledge building and sharing in software developer communities--allowing for an understanding of how software is built and why certain tools and features are adopted over others.- Led qualitative, quantitative, and mixed-method research: virtual ethnographic studies, case studies, interviews, surveys, focus groups, and data mining methods- Collaborated with experts and international software engineering researchers- Communicated our research through academic papers, blog posts, and by giving talks- Mentored, managed, and supervised other graduate students in their research

Designed and developed a code recommendation prototype tool (exampleoverflow.net) that uses questions and answers from Stack Overflow to help developers find and use high quality existing code snippets. I surveyed developers to inform the design decisions for Example Overflow, and conducted lab experiments and usability testing to evaluate the tool. This tool helped me investigate opportunistic software development.- Leveraged large data sets, web development, search algorithm development, human-computer interaction design, and more- Research methods: lab experiments, think-aloud protocol, a survey, observations, and interviews- Technologies: Java, jQuery, SQL, CSS, HTML, and Google App Engine