A Python pipeline (OpenCV, PyMuPDF, pandas) that turns scanned post-lesson reflections into structured data, so students keep pen and paper practice without the need for manual data-entry.
Being used to measure engagement metrics for an ongoing study of approximately 500 English learners across two high-schools in rural Japan.
Tracking engagement of ~500 high school English learners across two schools using worksheet-reader data and mixed-effects models (statsmodels).
Findings are being shared with JET Program supervisors to improve teaching methods and track student engagement, progress, and more.
Data is being collected and analyzed with the full permission of teachers and school administrators. Only anonymized statistics and plots will be shared publicly.
Act as the primary English instructor (T1) in 10-20 English classes per week at two rural Japanese high schools.
Work and live primarily in Japanese: about 50-60% of classroom instruction is delivered in Japanese, and Japanese is the main language used with colleagues.
Plan and lead lessons, activities and communicative exercises; create assignments and give evaluations.
Adapt to diverse learning needs, including neurodiverse (ADHD, ASD, dyslexia) learners.
Coach students for speech contests, debates, and research presentations.
Design after-school events that promote cultural exchange and language learning.
Collaborate with Japanese teachers on materials, curriculum and student engagement, acting as both educator and cultural ambassador.
Develop curricula and syllabi across math (2nd-12th grade and college level), computer science, machine learning, physics and SAT/ACT prep.
Tutor both students who are struggling and those advancing beyond their school's offerings.
Design, administer, and grade assignments and tests to assess progress and understanding.
Adapt instruction for a diverse range of students, including neurodiverse and non-native English-speaking learners.
Collaborate with parents, guardians, and fellow tutors on progress updates and concerns.
Used Excel, Microsoft Dynamics and Python to access, clean, and analyze business data for the M/WBE (Minority and Women-Owned Business Enterprises) program.
Geocoded messy, un-standardized data for thousands of businesses with QGIS and the Google Geocoding API.
Built visualizations and reports in Excel, Power BI, and Python that NYC SBS officials and local politicians used to monitor program success.
Completed mediation training in communication and conflict resolution.
Coordinated with colleagues across experiments and teams to schedule, then independently ran, daily test runs on the ATLAS New Small Wheel (NSW) in preparation for Run 3.
Analyzed and visualized data from ATLAS New Small Wheel (NSW) test runs with PyROOT, NumPy, SciPy and Matplotlib.
Compiled test-run data and error files into reports and visualizations of the NSW's operational status over two months, informing labor reallocation and funding decisions across teams.
Attended various lectures, workshops, and weekly meetings on various topics in high-energy physics.
Planned and executed a 12-week full-stack project delivering three end-to-end features on an internal Chrome Web Store platform.
Used Kotlin and protocol buffers for data storage, remote procedure calls and database queries.
Built with Guice (dependency injection), gRPC, the Wiz frontend framework and HTML templating libraries.
Read, wrote and maintained design documents
Read and edited other engineers' code in a large codebase.
Shipped production code through code review and supervisor feedback.
Attended several workshops and lectures on various topics in computer science.
Studied the basics of quantum mechanics, quantum computation and algorithms with MIT and Oxford professors.
Intensive four-week introduction to computer science for rising college freshmen, especially students from groups historically underrepresented in the field.
Capstone: a three-day JavaScript group project, a visualizer that generates graphics for music.