Hello!
First-generation student exploring software, hardware, and human-centered technology.
My name is Elvis Amaya, and I am a first-generation student with a passion to explore, intersect, and create.
I strongly believe in the strength of interdisciplinary thinking in creating effective human-centered solutions, and I am always eager to learn from different perspectives and experiences.
My ultimate goal is to use technology to erase the gap that currently exists between low-income students and their engagement in education.
In pursuit of this, I am exploring technology as a catalyst for creating passionate students who have an ever-expanding hunger to know more.
An interactive, story-driven web app that teaches machine learning fundamentals through a retro space adventure. Users travel across six unique "worlds" — from data collection to model evaluation — guided by Captain Nova, completing quizzes to unlock each stage.
Built with React and Vite. Features an 8-bit pixel art UI, XP tracking, dynamic planet visuals, and a world-progression system that locks stages until the user passes each mission.
View on GitHubAn interactive stock market simulator that teaches investing fundamentals through real-time trading, simulated market dynamics, and guided learning. Users practice buying and selling stocks in a risk-free environment designed to build financial intuition.
Built entirely in HTML with no external dependencies, making it lightweight and instantly runnable in any browser.
View on GitHubA lightweight convolutional neural network trained from scratch on the CIFAR-10 dataset, achieving solid validation accuracy in just 15 epochs. The model applies data augmentation, batch normalization, and dropout to improve generalization on unseen images.
Implemented in Python using PyTorch, with a focus on understanding how CNNs learn spatial features without relying on pre-trained weights.
View on GitHubA character-level transformer built from scratch in PyTorch to understand how large language models learn to generate text. The model learns token relationships through self-attention, covering the full pipeline from tokenization and embedding to multi-head attention and autoregressive text generation.
The goal was a deep, hands-on understanding of transformer architecture rather than relying on existing libraries or APIs.
View on GitHubA fashion-focused app prototype that helps users track their closet, build outfits, get style inspiration, and explore outfit posts. The app features a wardrobe library, an AI-powered style prompt generator, a social explore feed, and a profile with past fits — all running locally in the browser with no backend.
Built as a single-page HTML app with a clean, mobile-first UI. Includes closet stats, outfit shuffling, a post composer, and a mood-based inspo generator.
View on GitHubThis interactive Python project uses the turtle graphics module to simulate a colorful racing game between multiple turtles. Players choose the number of racers from 2 to 10, and each turtle moves forward by random distances as it competes to reach the finish line first.
Randomly placed obstacles on the track slow down or stop racers, adding an element of unpredictability and strategy.
This visualization compares Breadth-First Search and Depth-First Search in maze traversal. Each algorithm explores the maze differently, with BFS finding the shortest path layer by layer while DFS dives deep before backtracking.
Both are implemented in Python and animated using the `curses` library.
A contactless heart rate and respiration monitoring system built with a 77 GHz FMCW mmWave radar (TI IWR6843ISK). Instead of wearable or contact sensors like PPG, it measures sub-millimeter chest displacement from the phase of the radar signal.
I derived and implemented the full signal processing pipeline (range-FFT, DC offset correction, phase extraction and unwrapping, and slow-time FFT) and designed the chirp parameters to reach under 0.1 Hz frequency resolution, enough to separate breathing (0.2–0.5 Hz) from heart rate (1–1.6 Hz). Across multiple trials, the system reliably detected respiratory rate (14–19 bpm) and heart rate (61–80 bpm) from 1 m away. Stanford University, May–June 2026.
An autonomous mini forklift inspired by the Butter Robot from Rick and Morty. The robot uses a camera to find a stick of butter, slides its forks under a small pallet, lifts it, and delivers it to a red drop-off marker.
An STM32 Nucleo-L4R5ZI-P runs the robot as an interrupt-driven state machine. It receives butter and marker positions from a XIAO ESP32S3 Sense camera over UART, reads heading from an LSM303AGR magnetometer over I2C, and drives two DC gearmotors and an SG92R servo with timer PWM. The 25-part 3D-printed body includes a parallel-link forklift that lifts the forks 30 mm while keeping them level. In progress, Fall 2026.
This project features an Arduino Uno board along with various electronic components to power the feeder system. It uses a motion sensor to detect an animal's presence and activates a motor that releases food.
By providing a reliable food source, the feeder aims to reduce the risk of species decline and support the balance of ecosystems.
I designed and built a solar-powered USB charger capable of charging small electronic devices such as cellphones, iPods, and tablets. The system uses a solar panel to capture sunlight and convert it into electrical energy, which is then stored in a rechargeable lithium-polymer battery.
Because USB devices require a constant 5V output and the LiPo battery provides only about 3.7V, the build integrates a DC-DC voltage converter to step up the voltage.
This project involved designing and building a motorized "useless box," a playful gadget that turns itself off whenever switched on. When the user flips the external toggle switch to the on position, a mechanical finger driven by a DC motor emerges from the box, flips the switch back off, and then retracts inside.
The system uses three AA batteries, a DPDT switch for directional control, and an SPDT microswitch to detect when the finger is fully retracted and stop the motor.
A multitasking embedded firmware system built on an STM32 Nucleo board using FreeRTOS. The system runs concurrent tasks for sensor polling, motor control, status LEDs, and a UART command shell — all coordinated through task priorities, preemption, queue-based messaging, and mutex-protected output.
Commands like motor open, led on, and
status can be issued live over serial. Hardware
includes the STM32 NUCLEO-F446RE, an SG90 servo, and an external
5V supply. Firmware written in C.
As a part of the Networked Systems Lab in USC's Viterbi School of Engineering, I collaborated with professor Ramesh Govindan and Ph.D. student Rajrup Ghosh to explore the intersection of technology, history, and education.
By using Unity, an Oculus virtual reality headset, and a computer, I constructed historical scenes within virtual reality so teachers and students could immerse themselves in the history of the U.S.
If you'd like to connect, collaborate, or learn more about my work, you can reach me through email, GitHub, or LinkedIn.
© 2026 Elvis Amaya. All rights reserved.