About
Computer Science @ UCLA
Hey there! I'm Brandan, a CS student from Southern California. In my freetime, I enjoy maintaining
a consistent weightlifting routine and spending time outdoors playing soccer or table tennis.
This year, I’m interning at Cedars-Sinai, where I embedded and networked a wearable medical device
for reflex testing, and deployed a companion app used by clinicians to pair with this device.
I am currently a student researcher in the System Security Architecture lab at UCLA (SsysArch),
I work on projects that protect privacy and make computer vision and wireless systems more reliable.
Resume
Education & professional experience
Education
University of California, Los Angeles
Computer Science student graduating in 2027 with a focus on Machine Learning and System Security
- (949) 524-6525
- brandansb@ucla.edu
Professional Experience
Software Engineer Intern - Mobile Dev & Networking
Mar 2025 – Present
Cedars-Sinai - Surgical Innovation and Engineering, Los Angeles, CA
- Developed control software in C for motor reflex wearable neurostimulation devices, implementing TCP socket communication and BLE protocols to transmit sensor data from Arduino hardware to backend.
- Designed and deployed networked mobile application in React Native with FastAPI backend for clinicians, enabling secure real-time data streaming from wearables to InfluxDB and AWS RDS.
- Architected cloud infrastructure on AWS (Elastic Beanstalk, RDS, S3) to host backend APIs, ensuring scalable, low-latency device-to-cloud communication for live signal monitoring.
Software Engineer Intern - Backend
Jun 2024 - Sep 2024
Hotb Software, Irvine, CA
- Designed 32 REST API endpoints in Express Node.js for financial loan assurance platform’s admin portal
- Implemented core administrative logic including CRUD operations for user loan application, role-based permission controls, and audit logging for data/files
Research
LightPure: Real-Time Adversarial Image Purification
Dec 2023 – Sep 2024
Research Assistant, ML Researcher — SsysArch Lab, UCLA
- Co-authored ACM MobiCom ’24 paper on a two-step diffusion + one-shot Generative AI framework for adversarial purification on Jetson Nano.
- Built and trained lightweight neural network achieving 10× faster and 2× more efficient image cleanup compared to prior methods.
Cloak of Invisibility: Privacy-Preserving Video Streaming
Apr 2024 – May 2025
- Created real-time video pipeline that automatically blurs or hides private objects (people, screens) in live RGB-D camera feeds.
- Integrated object detection with depth checks, achieving 88% accuracy at >30 FPS and SSIM ≈ 0.98.
Janus: Radio Frequency Spoofing
Jan 2025 – May 2025
- Developed signal generation model for the first over-the-air adversarial attack against Bluetooth device fingerprinting AI using SDRs.
- Designed signal tweaks causing >95% misidentification at 2m and >60% at 6m, viable through walls.