Mordecai Araya

▣

About Me

I'm an undergraduate Computer Engineering major at UC Santa Barbara (from the Bay Area), interested in Hardware, Robotics, and Machine Learning research. Whether it's tinkering with robotics or researching edge devices: I'm always driven to explore and build greater.

This summer, I've been researching neuro-inspired computer architecture for energy efficient bionic vision applications. This entailed building truly sparse, event-based representations for real-time, low-power retinal prosthesis stimulation on edge hardware like the NVIDIA Jetson Orin Nano.

Outside of research, I've built physical systems to demonstrate my passions for the tangible interface between hardware, AI, and real-world application. Case in point, building a on-person prototype for a bionic vision processing system as well as fully designing, modeling, and building an award-winning robotic arm from scratch.

▣

Research

Research Assistant, Arch Lab

Implementing specialized computing architecture for temporal-computing models, to support real-time and efficient processing for bionic vision research efforts.
[Embedded Systems, Linux, C++, CUDA]

AI-Brain Alignment Summer Research

Developed methodology to evaluate neural activation to streamline fMRI & EEG data by over 50% while simultaneously maintaining image reconstruction performance. Concluded program by presenting our work and results for the Summer Institute of Science and Mathematics program.
[Deep Learning, Mindeye2, Diffusion, Python]
▣

Projects

ARM-01: Robotic Arm

Fully functional 6-servo robotic arm designed and built from the ground up.

[Arduino, Servos, C++]
Gaucho Guardian: Course Scheduler

A web application for managing and scheduling courses at UC Santa Barbara (SB Hacks XII).

[Python Flask, APIs, HTML, JSON]
Autonomous Robotic Car

Real-time object detection using sensors, logic, and arithmetic. Developed from scratch using VEX Chassis.

[C++, VEX V5 Platform, Analog & Digital Sensors]
Facial Emotion Detection Algorithm

Trained an AI to classify human emotion in images in 1 of 7 categories, employing edge detection and Euclidian distance to deliver an 80% accuracy across 1,000 images.

[Python, TensorFlow, NumPy]
Photography & Editing

Freelancing, Photo & Video editing, and Social Media.

[Lightroom, Premiere Pro, Figma, Canva]
▣

Contact

If you have any inquiries, feel free to reach out at mordecaiaraya07@gmail.com.