2025 - Formula SAE Car

A rendered version of UCLA's MK10 racing car made in Blender. Adopting Blender improved the design and manufacturing of the livery. From a design standpoint, it allowed us to better understand how material and color choices would look in a photorealistic environment. From a manufacturing standpoint, we could use the textured model for exact size and placement reference on the real car. These improvements gave us the confidence to pursue a lower contrast look and use a mix of automotive spray paint and vinyl.

Responsible for all texturing, simulation, and compositing of the renders as well as manufacturing the car's livery.

1. X-Ray Assembly

2. Indoor Turnable

3. Wind Tunnel

4. Static Renders

2024 - Space Opera

This space epic is a compilation of various intergalactic phenomena. The project served as an avenue into advanced procedural and atmospheric shading. All renders were made in Blender and edited in DaVinci Resolve.

1. Nebula

Responsible for the procedural creation of the nebula geometry and shading.

2. Sun

Responsible for the procedural sun shader.

3. Black Hole

Responsible for the procedural creation of the accreation disk shader.

4. Earth

Responsible for the volumetric rendering of the atmosphere. Earth scans are taken from NASA.

2024 - Sal

Sal is an on device iOS chatbot app written from scratch using Swift. Using llama, whipser, and espeak, it carries out basic conversations without relying on APIs from OpenAI, Google, Meta, and the like. To look at the code, visit the github repo. Check out this blog for deep dives into its construction. Responsible for the programming and design.

2025 - Knot a Film (IN PROGRESS)

Knot a Film is a visualization and recommendation tool for movies. Using Kaggle's rotten tomatoes dataset, over 60000 movies are parsed into an SQL database of movies and their associated actors, directors, and writers. Once this representation is achieved, a force node runloop simulates the positions of nodes based on these connections for the final output. In addition to this visualization, another part of application uses custom word embeddings to query the dataset for relevant movies. The code in its current state can be viewed here. Responsible for the programming and design.