Automated a role request system by building a REST API with HTTP routes that leverages LDAP queries. Streamlined notifications via a Slack bot designed on Bolt that formats role requests and regularly pings theoperational status of applications. Developed a scalable user directory for over 3000 users via API pagination for better performance and query parameters for advanced search capabilities. Coordinated the migration of a flagship application to a new OS version, covering MongoDB database cleanup and production environment setup.

Upgraded Fintech UI framework of 200+ components to modern React standards through functional component conversion and React Testing Library usage. Improved build-time quality assurance by enforcing type checking and raising compile-time errors with TypeScript. Increased Playwright test coverage by 20% via E2E testing for future bug detection and enhanced system stability. Engineered a Python client-friendly API for 30+ page object models to maximize ease of authoring and maintenance. Collaborated in daily scrum meetings and conducted best agile development practices: planning sprints, resolving blockers, and refining issues. Showcased internship project to stakeholders, cross-functional teams, and senior managers (including the CEO) as part of a company-wide update.

Validated the performance of both established and proprietary machine learning algorithms (FedAvg, FedProx, RGT-FedAvg) on 5 heterogeneous datasets to prove the latter performed best. Implemented theoretical federated learning algorithms with NumPy and pandas, including the local training (stochastic gradient descent), gradient tracker, and model aggregation. Analyzed methods on MNIST dataset image recognition, translating multidimensional matrices in binary classification. Visualized loss function value vs. number of epochs passed utilizing plots generated by Matplotlib. Reported progress updates with the team weekly to present in front of 500+ faculty, staff, and peers at the Undergraduate Research Symposium.

Taught coding classes (Python, Java) to students ages 10-14 and organized the curriculum for each course. Brainstormed and designed coding projects focused on Big O notation, recursion, data structures (linked lists, hash table, binary search trees), and other topics. Spearheaded initiatives to proactively optimize 24/7 server communications for a community of 1000+ members across 20+ states

Implemented a maximum likelihood algorithm to classify flooded regions in New Jersey caused by Hurricane Ida, enabling proper budget allocation for preventative measures. Presented my findings and interpretations to the environmental sensing team biweekly so that an AI-powered application can be cross-validated

Resolved 50+ mice movement database errors, which refined the publication, ”The Behavior Biomarker Scale (BBS): A machine-vision approach for automated locomotor recovery evaluation at millisecond timescales”. Gained a certificate that grants the status of co-authorship for the publication

Coordinated networking events, workshops, and judging process in collaboration with a cross functional team to ensure successful execution of a hackathon. Streamlined the mentorship of 20+ teachers despite pandemic limitations by dictating workflow