Hey!
Thanks for visiting my website -- more updates are to come! My name is Zack, and I'm a fourth year Computer Science and Data Science student at UC Berkeley. I love looking into new topics, and I'm currently doing computer graphics, computer vision, and full stack development.This past summer, I worked as a SWE intern at Amazon, designing and implementing a data pipeline, intermediate database, website backend, and a frontend dashboard that ingested 600 million records of Amazon.com pricing data by the end of my internship.
Before that, I worked as a SWE intern at Space of Mind (mental health startup) implementing a prototype of their patient chatbot, semantically chunking their transcript data into a vector database, and using LangChain to provide query context for GPT-3.5.
What I'm currently building: An (eventually) open-source online visualizer for vision scientists to visualize 4D (or even N-D) object color solids.
In order of recency, some projects I've worked on:
SPLAT! (link)
Developed on the Snapchat Spectacles, you color your room with your voice! The pitch of your voice determines the color of the paint. Splats are both voice and hand triggered. Done in Typescript on Lens Studio.
Procedural Terrain Website (link)
Using OpenGL (in C++) and shaders, I implemented moving water in the procedural terrain, complete with reflection and diffraction. Read more about the methodology in the link above (under the "Final" tab with the "Water" heading).
Face Morphing
|
|
Using Delaunay triangulation, affine transforms, and a cross-dissolve, we can achieve a fun face morph effect!
Technologies: Python, Numpy, SciPy, OpenCV
PintOS, Building Functionality Into an Operating System
Implemented (in C):
- User programs: terminal argument passing, process control syscalls, and file operation syscalls
- Threads: multithreading, strict priority scheduling, alarms
- Filesystem: buffer cache implementation, an extensible file system similar to FFS, and subdirectory traversal/path resolution
Let Them Cook (link to description)
This application recommends recipes based on images of food in your fridge.
Technologies: React, TypeScript, Flask, SQLite, Tailwind CSS, Python, LangChain, Pinecone
DittyCal (this one is deployed!)
Here, we recommend music based on left and right swipes, like Tinder.
Technologies: React, JavaScript, Flask, Python, Pandas, NumPy, Scikit-Learn, Spotipy, HTML, CSS
Addendum:
You can also find my Linkedin here, and my Github here.What I'm currently learning: Volleyball, Blender, Three.js, and Mandarin!
(Note, to go back, you'll need to click the laptop again)
You can (after leaving this overlay):
- rotate the camera by dragging your mouse
- pan the camera by pressing shift and dragging your mouse
- zoom in/out by scrolling up/down
I'll be moving this website to MantineUI, with R3F, as it currently isn't mobile friendly.