Experience

At PNI Sensor, I design and implement advanced sensor fusion systems that power mission-critical applications in GPS-denied environments. I built a C++ middleware that reduced integration time across platforms by 50x, developed Rust pipelines that ingest millions of data points for model training, and created robust serial/BLE encoders that cut packet loss to under 0.1%. My work spans across several platforms, delivering reliable positioning and navigation solutions to real-world defense and commercial use cases.

I joined Davis Housing Services, a student-focused startup, to build secure and scalable backend systems. I implemented Firebase authentication with JWT-based sessions, designed REST APIs with FastAPI and MongoDB, and introduced asynchronous data handling for efficiency. To improve matching between exchange students and host families, I architected an LLM-driven algorithm that quantified subjective preferences like personality and lifestyle, creating a more human-centric and accurate experience.

At C-Square Lab, I researched the sustainability of open-source communities by analyzing governance policies across large repositories. I built ETL pipelines and clustering algorithms that visualized how new contributors emerge and interact, and I designed validation methods to test LLM performance on complex NLP tasks such as co-reference resolution and semantic role labeling. This work combined technical rigor with social science insights to study the dynamics of digital communities.

At UC Davis, I taught and mentored students in a three-unit competitive programming and advanced algorithms course. I developed problem sets, guided students through concepts like graph traversal, dynamic programming, and complexity analysis, and emphasized practical strategies for coding interviews and competitions. Beyond teaching technical material, I helped students build confidence in approaching unfamiliar problems, fostering a culture of curiosity and persistence in problem-solving. Additionally I had to develop automated testing for student code submission and assignments through the Gradescope API.

I contributed to firmware development for ARM-based SSD controllers, focusing on both performance and reliability. I implemented disk write policies that reduced write times by 12% while safeguarding against data loss, and automated over 50 regression tests to stress-test firmware under demanding conditions. By migrating 100+ tests from Python 2.7 to Python 3, I cut regression time from 73 hours to 42, significantly accelerating the development cycle.

At NOMAD, a property rental startup later acquired by Wish STR Enterprises, I worked on building core backend infrastructure. I created a database seeding pipeline to simulate load, developed geospatial caching that doubled indexing efficiency and reduced query latency, and helped orchestrate microservices deployed through Kubernetes on AWS and Firebase. My contributions supported NOMAD’s rapid growth, including winning $50,000 through the UCLA Venture Accelerator.