Since August 2024, I have served as a Research Assistant at Tulane University under the guidance of Dr. Aron Culotta, working on research focused on applying machine learning methods to the legal domain. My work centers on developing deep learning models to analyze administrative datasets and predict procedural outcomes. As this research is currently unpublished, I must remain somewhat general in describing the specific methods and applications. I built the entire pipeline from the ground up, including data preprocessing, feature engineering, model development, and evaluation infrastructure. This project has deepened my expertise in neural network architectures, model interpretability, and the responsible application of machine learning to sensitive contexts, while reinforcing my commitment to careful consideration of the ethical implications inherent in deploying AI systems within high-stakes domains.

During my time as a Research Assistant at the University of North Texas, I immersed myself in a deep learning project under the guidance of Professor Ting Xiao and Dr. Daniel Lersch. My involvement originated from the Research Experience for Undergraduates (REU) program, focusing on the application of variational autoencoders for particle identification within the realm of high-energy physics. As part of the research team, I contributed to the collaborative design and implementation of the models. My individual contributions included building an automated hyperparameter tuning pipeline that reduced model selection time from days to hours, developing an analytics system for learning-curve analysis, anomaly detection, and latent-space evaluation with real-time monitoring, and implementing custom optimization routines to automate model assessment and selection. I presented this work as a poster at the REU symposium, where we reported empirical results and discussed methodological limitations and future directions. Beyond the REU, there was an opportunity to extend this research into a longer-term project. Over the subsequent time, I continued to refine and expand upon our initial findings, that lead to significant progress in model performance.

In the summer of 2023, as a Research Assistant at Purdue University under Professors Daniel Schiff and Kaylyn Jackson Schiff, I embarked on an exploratory project to harness the potential of Large Language Models (LLMs) for the analysis of sociopolitical behavior from a demographic perspective. My role entailed a thorough examination of LLM architectures to enhance their precision in identifying demographic-specific linguistic patterns. I crafted innovative prompts that married linguistic insights with AI technology, aiming to draw out nuanced information from varied demographic segments. Analyzing the predictive accuracy of LLMs in pinpointing distinctive phrases within open-ended survey responses, I applied advanced natural language processing techniques to large datasets to uncover the complex relationship between language usage and demographic factors. This multifaceted project not only honed my technical and analytical acumen but also deepened my understanding of the ethical dimensions of synthetic data generation. The experience solidified my resolve to advocate for AI's responsible use, recognizing its power and the necessity of thoughtful application.

As a tutor for CMPS 6620: Artificial Intelligence, a graduate-level course taught by Professor Aron Culotta, my role was to support M.S. and Ph.D. students as they developed both theoretical and practical foundations in AI. The course focused on the design and analysis of autonomous intelligent agents, combining formal algorithmic reasoning with implementation-driven assignments. I held weekly online office hours where I worked one-on-one with students to clarify core concepts such as uninformed and heuristic search, adversarial search, Bayesian networks, decision networks, Markov Decision Processes, and reinforcement learning. Much of my support centered on helping students reason through algorithm design choices, analyze correctness and complexity, and translate mathematical formulations into working implementations for programming assignments. I also provided targeted guidance on long-form assignments and exam preparation, helping students connect lecture material, textbook readings, and their own code. This role required communicating complex ideas clearly to a diverse group of graduate students and reinforcing rigorous problem-solving habits expected in an advanced AI curriculum.

As a teaching assistant for CMPS 1600: Introduction to Computer Science II, the second course in Tulane’s introductory computer science sequence taught by Professor Alireza Shirvani, I supported undergraduate students as they learned structured software design and data-structure–based thinking. I led weekly recitation sessions for two sections totaling approximately 60 students, where recitations were run as hands-on lab periods focused on weekly programming assignments. During these sessions, I helped students debug code, review implementations, reason about memory management and pointers, and analyze time complexity while reinforcing core object-oriented concepts such as inheritance, abstraction, and interfaces. I also held weekly office hours providing individualized support on assignments, exams, and the semester-long class project, where students built simple games using Professor Shirvani’s Camelot virtual environment. The project was structured around graphs, with nodes representing story states and edges representing player choices, and I assisted students with implementing this structure in Java, integrating it with the Camelot API, and testing their projects end to end. In addition to instructional support, I proctored the midterm exam. I also developed supplementary practice problems and review materials to reinforce lecture content and common implementation issues.

As a teaching assistant for CMPS 3160/6160: Introduction to Data Science, an undergraduate and graduate cross-listed course taught by Professor Rebecca Faust, I supported students as they developed practical skills in data analysis, statistics, and data-driven problem solving using Python. I led weekly lab sessions for approximately 40 students, where labs were structured as hands-on, applied work with real datasets using tools such as NumPy, pandas, and visualization libraries. During these sessions, I provided real-time debugging support, helped students reason through statistical concepts, and guided them in translating analytical goals into working code. I also held weekly office hours offering individualized assistance on labs, exams, and the multi-stage course project, and I organized and led a final exam review session to help students consolidate key concepts and prepare for cumulative assessment. In addition to instructional support, I developed detailed grading rubrics in Gradescope and individually reviewed each student’s submission, carefully evaluating code and written responses manually due to the limited automation available in the grading platform. I also helped create a practice midterm exam to allow students to assess their readiness and identify gaps in understanding.

As a mentor for STA-230, my role was to support and enhance the educational experience of students enrolled in this data science program. The course, who was taught by Professor Ryan Miller, was structured as a series of workshop-style classes, where the majority of the time was dedicated to hands-on lab activities, allowing students to collaboratively engage with the material. During this period, I guided students during bi-weekly class sessions focused on R programming and data analysis. These sessions were designed to deepen students' understanding of data science principles and to foster active participation. I also conducted weekly mentor sessions, during which I provided further explanations of course concepts and assisted with inquiries about homework and class materials. My responsibilities included efficiently managing communications, where I ensured that all queries from students and professors regarding course materials were promptly and clearly addressed. The curriculum I assisted with covered a range of topics crucial to data science, including data visualization principles and techniques using ggplot2, data manipulation with packages such as tidyr and dplyr, and the development of interactive web applications with tools like plotly and R Shiny. We also explored topics like principal component analysis and clustering, and applied various modeling techniques for both numerical and categorical outcomes. This mentorship position allowed me to not only apply my knowledge of data science to assist students but also to refine my communication and leadership skills within an academic setting.

In my role as a grader for STA-230, I was integral to the academic evaluation process, engaging in regular grading meetings to assess the work of a diverse cohort of 50 students. My commitment to upholding high grading standards was a cornerstone of this position. I collaborated closely with the course instructor to navigate the complexities of grading, especially when it came to ambiguous cases, ensuring that each student received a fair and balanced assessment of their work. My responsibilities extended beyond grading; I provided students with insightful feedback on their assignments, emphasizing areas in need of improvement as well as acknowledging their mastery of key concepts. This feedback was tailored to support and enhance the students' learning journey, helping them to identify their strengths and areas for growth within the data science discipline. This role underscored my dedication to academic integrity and my ability to contribute constructively to the educational development of students.

As a grader for CSC151, I was deeply involved in the foundational computer science course in the curriculum. The course is known for introducing students to fundamental computer science concepts such as recursion, abstraction, scope and binding, modularity, and algorithm design and analysis using Racket, a high-level functional programming language that is a variant of Scheme. In my role, I assisted in grading a range of introductory computer science projects and homework. The grading sessions I participated in were aimed at providing consistent and fair evaluations of the students’ coding assignments each week. Furthermore, I coordinated with the professor to develop effective grading methodologies, which was crucial for maintaining the integrity of the course’s rigorous standards. This engagement with CSC-151 allowed me to contribute significantly to the academic development of students by providing them with constructive feedback on their algorithmic problem-solving skills and programming techniques. My role also gave me a deeper insight into the pedagogical approaches necessary for teaching complex computational concepts​​​​​​.

As President of the Computer Science Graduate Student Association at Tulane University, I lead the elected graduate student governance body representing all of the CS graduate students (Master's and PhD) within the Department of Computer Science. I oversee the planning and execution of professional development events, research seminars, and social activities designed to foster community and enhance the graduate student experience, and act as the primary point of contact between CS graduate students and departmental leadership. In this role, I work directly with faculty and administrators on issues related to graduate student welfare, funding priorities, and access to departmental resources.

As a Computer Science Department representative to Tulane University’s Graduate Studies Student Association, I advocate for the interests of Computer Science graduate students within the university’s officially recognized governing body for master’s and doctoral students in the Schools of Liberal Arts and Science and Engineering. In this role, I collaborate with representatives from other departments to review and vote on funding requests and policy proposals, helping allocate student activity funding for academic, career, and social initiatives that shape graduate student programming and resource distribution. Additionally, I serve on GSSA’s Legislative Committee, helping shape policy language and decisions that influence funding processes and support structures for graduate students across the schools.

As a Senator in Tulane University's Graduate and Professional Student Association, I represent the GSSA division within the university-wide governing body for all graduate and professional students. GAPSA is the unifying student government for all graduate and professional divisions, charged with advocating student interests to university administration, promoting cross-division communication, and coordinating programming that supports academic and professional growth across disciplines. In this role, I review funding proposals, participate in deliberations on policy matters that affect graduate students broadly, and vote on measures related to funding and student support. I work across divisions to ensure that shared concerns — such as interdisciplinary programming, equitable funding distribution, and resource access — are considered in GAPSA governance.

As a Co-Founder of the Machine Learning & AI Club at Grinnell College, I helped establish a student organization dedicated to building technical skills in machine learning and applied artificial intelligence. I organized and led hands-on workshops that taught students to build and evaluate models in areas like neural networks, natural language processing, and computer vision. I also coordinated speaker events with industry professionals to connect students with current practitioners and career pathways. As an alumni advisor, I continue to support club leadership by advising on curriculum planning, technical direction, and maintaining industry relationships that help the club offer relevant, practical experiences for members.