Killing the "where should we eat" debate with targeted filtering

OffMyMenu grew out of a personal running list of food spots I'd been building since college. The goal was simple: kill the endless "I don't know what to eat" back-and-forth by giving people a fast, low-effort way to land on a place to grab food. I built it so users can narrow down the collection based on whatever filter they need, like cuisine or location, and included a random selector (which also uses those filters) for when you just want the app to make the decision for you completely.

On the design side, I jumped into Illustrator to build out a visual identity from scratch, leaning heavily into a scrapbooking and journaling vibe tied together with a Polaroid photograph motif. I also integrated the Google Maps API so actually navigating to the spots was frictionless, a feature I prioritized since most of the college students I knew were browsing on their PCs. It was one of my first real ventures into full-stack thinking, balancing a custom frontend aesthetic with actual data usability.

Developing a Secure Password System with Accelerometer-Based Authentication

I collaborated two others to develop a secure and innovative password system for our Embedded Systems Design class competition. The core of our project was utilizing the ATmega32U4 processor and Arduino library to create an accelerometer-based authentication method. This involved recording a hand movement sequence as a unique key to unlock resources, deviating from traditional text-based passwords. Our approach involved low-pass filtering functions to smooth accelerometer data, ensuring accurate motion detection. The recorded data was then compared to attempted passwords, determining success based on predefined tolerance levels.

The hardware configuration of our project included the Circuit Playground Classic Board which had an built-in accelerometer for motion data capture. Data handling was managed through functions like lowPassX(), lowPassY(), and lowPassZ(), recording data into arrays for saved and attempted passwords. Performance optimization was achieved by considering the use of pointers for memory efficiency. This project for me emphasized the importance of collaboration, showcasing how diverse expertise can lead to comprehensive solutions.

Integrating Systems for Autonomous Underwater Vehicle (AUV) Design

As a Computer Engineering major and member of the Operations division within the AUViolets, NYU's RoboSub team, I contributed extensively to the development of our robotic submarine designed for the international RoboSub competition. This project required integrating mechanical, electrical, and software engineering principles to create an Autonomous Underwater Vehicle (AUV) capable of completing complex tasks with computer vision and sensor guidance. The AUV utilized the NVIDIA Jetson Xavier NX CPU as its processing unit and incorporated over 25 on-board sensors for navigation and task execution.

During the project, I was responsible for integrating over 25 on-board sensors with the NVIDIA Jetson Xavier NX CPU, ensuring seamless data flow and interaction between components. I developed and refined detailed wiring diagrams using Autodesk Eagle, focusing on quality assurance of power management and adherence to hardware schematics. This included verifying the electrical load distribution to prevent overloading and optimizing sensor placement for accurate data collection. I collaborated with a team of 5 members within the Operations division, serving as the intermediary between electrical engineering and computer science teams to maintain alignment between hardware and software functionalities.