Let's build the future
of technology

A current summer 2026 intern, working in a 4-person machine learning and computer vision team. I have the sole ownership of enhancing the inference and testing Python-based toolchain to evaluate the performance of Hover's two computer vision models: YOLO and RF-DETR. Through this pipeline, I've been able to identify which model is performing better on unseen data. These findings serve as key benchmarks for how the next training jobs are run for our computer vision models.
Throughout this internship, I've gained invaluable fluency in cloud engineering with Amazon Web Services' (AWS) SageMaker AI, migrating the entire inference and testing pipeline onto the cloud through AWS processing jobs. Alongside this, I've gained experience with Docker Desktop, WSL, Ubuntu, AWS, Git, utilizing Gemini AI, and VSCode across all my coding projects.
Next, I'll be working on developing the writing portion of a SQL database to streamline the generation of training data for training our machine learning computer vision models.

Explored soliton behavior in fiber optic rings through the Lugiato-Lefever Equation (LLE), implementing pyLLE, an open source LLE solver, to simulate frequency comb behavior in micro resonators. Generated normalized intensity and spectral intensity visualizations to characterize soliton drift and behavior over time. Earlier work in the lab comprised collecting spectra and mapping data for 12 peptides using a Raman spectrometer, specifically an Invia microscope, in partnership with startup Primary Bioscience.

Developing a compact Detect-and-Avoid system using camera imaging and object identification. Participated in initial paper research and project outline development, as well as the training of YOLO models for real-time aerial object detection. More details on the project can be found on the project website.

Designed, manufactured, and assembled the chassis for a Mars rover competing in the International Rover Challenge Series. Familiarity with machining parts through bandsaw, mill, 3D printing, and rubber molding. Did motor and gearbox calculations for the 2026 competing rover. Also served as the primary liaison between mobility and electrical subteams, coordinating motor communication, encoder selection, and electronics placement.

Utilized Tableau and Python to analyze 2000+ Kode With Klossy scholar final projects, identifying trends amongst curricula comprehension and rising popular final project topics amongst the youth, such as mental health and climate change. Also built a Census/Labor Bureau API project that collected data on electricity access and water access to identify optimal U.S. semiconductor fabrication plant locations. This project served as a example project for an Infosys challenge run in partnership with the Kode With Klossy organization.

Taught introductory AI/ML to 13-18 year-olds over a 3 day workshop hosted at Amazon Headquarters in Seattle, Washington through the design of an image classifier.

Collaborated between a team of 7 to plan daily daily curricula to coordinate groups of 100+ scholars over zoom. Ensured scholar understanding of and proficiency in Python/SQL, facilitated remediation lessons, created “cheat sheets”/demo videos, and hosted office hours.