AI product, software, and research.

Built an automated mobile-app QA system using Python image processing and the Gemini API, then transferred the codebase to the deployment team for integration into Rudel's live QA workflow. Engineered a custom evaluation metric that penalizes false negatives to improve defect-detection reliability.

Led product strategy for an AI-powered B2B concept for Indeed's ATS, recognized as the program's most promising product. Synthesized 73 survey responses and 34 stakeholder interviews to drive a pivot, then built 80% of the JavaScript/Python prototype and delivered the concept within two weeks.

Built a high-fidelity Isaac Gym simulation of the Unitree G1 humanoid and integrated a vision-language model into the task-and-motion-planning stack to handle scenes beyond the prior rule-based pipeline.

Developed a facial-expression-recognition model for fisheye images by researching and implementing deep-learning architectures in PyTorch in collaboration with academic researchers.

Designed and executed 15+ experiments with two PhD researchers on diffusion-inspired transformer attention efficiency, characterizing the efficiency-accuracy tradeoff at longer context lengths and quantifying projected compute and memory savings to guide the research direction with the advising professor.

Built a C++/Unreal Engine visualization tool for NASA Johnson Space Center to simulate Lunar Gateway and surface operations. Directed integration of Canadarm3 robotic-arm kinematics, fulfilled all functional requirements, and delivered a live technical demonstration to NASA leadership and project stakeholders.

Revamped LaunchPad's website to drive greater community awareness and engagement. Collaborated in weekly strategy meetings to identify challenges and implement support initiatives for both mentors and mentees, enhancing overall program experience. Cultivated and sustained sponsor relationships through regular outreach and partnership updates, ensuring ongoing program funding and resource alignment.

Selected as 1 of 31 from 3,881 applicants (top 0.8%) for the August 2025 program. Interviewed 50+ residents and stakeholders, identified drivers of women's underrepresentation, and presented targeted solutions to the mayor. Invited back seven months later, authored a follow-up proposal, and returned on March 22, 2026, to present “Inubari.”

Founded an organization to promote sustainable living through technological initiatives. Programmed and launched a browser extension supporting Amazon in 4 countries, India, Japan, the UK, and the US, that calculates the CO2 footprint of products and shipping based on dimensions, weight, country of origin, and category.

Developed and conducted further research on Facial Emotion Recognition AI that interprets emotions from images at the University of Tsukuba.

Conducted archival research on AI model accuracies and algorithms for FER and presented findings at Chiba University. Selected as the best researcher and was sent to Chulalongkorn University, fully paid.

Selected as the top researcher by Chiba University and sponsored to conduct further research at Chulalongkorn University.

Organized and managed an online seminar series on Sustainable Development Goals awareness for over 40 selected students. Led the event team in identifying target participants, developing presentation slides, and overseeing all event-related activities. Collaborated with international teams from Canada, Indonesia, and the US to drive global engagement on sustainability topics.