Musab Fiqi

Spearheaded the development and maintenance of the center’s primary website, ensuring a user-friendly and informative online presence.

Engineered and implemented a robust technical infrastructure, including computer systems, network architecture, printers, and essential technology accessories, optimizing operational efficiency.
Designed and deployed classroom technology setups tailored to individual teacher needs, fostering an engaging and technologically enriched learning environment.

Leveraging TypeScript, I architected a WebGPU-powered front-end application to showcase complex data visualizations, including direct volume rendering and ray tracing of tetrahedral meshes.

Using the WebGPU Shading Language (WGSL), I implemented a sequential mesh traversal algorithm that accumulates color and alpha values to accurately visualize complex, unstructured data.
Achieved visualization of turbulent airflow around a golf ball.
Identified performance limitations due to CPU-intensive pre-processing steps; future work will focus on utilizing WebGPU compute shaders to offload these steps and enable real-time rendering of larger, more complex datasets.

WebGPU Typescript WGSL Real-time Rendering

Utilized PyTorch and Jetson Nanos to train a convolutional neural network for identifying various weed species. Jetson Nanos were attached to a stinger drone which was flown over crops to detect weeds such as canopy, waterhemp, giantragweed and marestail.

Made the process of extracting frames from the recorded videos multithreaded, which substantially increased the speed at which frames were extracted.
Processed 12TB of frames which helped increase the accuracy of the CNN model to 92%..

Python Pytorch Convolutional Neural Network Jetson Nano

Collaborated closely with a select team of undergraduate students to design and program a Sociology research study, funded by the esteemed National Science Foundation. Utilized JavaScript within the Lioness Labs tool to dynamically modify various study behaviors that was requested by the graduate research student.

JavaScript Lioness Labs

The Ohio State University

M.S. in Computer Science and Engineering - August 2023 - December 2024

Specialized in Computer Graphics.

Developed a 3D scene using physically-based rendering (PBRT), incorporating advanced rendering techniques like bump mapping and environment mapping for enhanced realism. The scene was rendered with 640 samples and a maximum depth of 160, resulting in high-quality images despite the challenges of complex materials and lighting.
Developed a WebGL renderer enabling real-time rendering in 3D with lighting, textures, and environment mapping. Proficiently utilized JavaScript and GLSL.

The Ohio State University

B.S. in Computer Science and Engineering - January 2020 - May 2023

Specialized in Artifical Intelligence.

Designed and implemented a Spotify playlist recommendation system utilizing Elasticsearch and React, where the user prompts the program to generate personalized playlist recommendations based on the input songs. Developed a specialized algorithm which considers various audio features such as danceability, energy, tempo, etc.
Developed an image classification system using a convolutional neural network. Classified images from the CIFAR-10 dataset using standard neural network techniques with an accuracy of 75%.

Columbus State Community College

Associate's of Science- January 2018 - December 2019

You may think you know the story of the ugly duckling, but you haven't seen it like this! The swan got tired of being bullied by all the other ducks, and now seeks revenge! Punch your way through these malicious mallards in search of your mother who disowned you! How will you treat her when you finally reunite? Well... that's up to you. Try out the game here!

Unity C#

Developed a cyberpunk-themed FPS game in Unity. Implemented gameplay mechanics such as combat/environmental interactions and player gunplay.

Imeplemented dynamic enemy AI with intelligent aiming and shooting mechanics, enabling varied combat encounters based on player distance and unique flanking movement patterns.

Try out the game here!

Unity C#

Recreated the first dungeon in The Legend of Zelda (1986) with some added rooms inspired by Doom (1993). Implemented common software design patterns such as the singleton, factory, decorator, and state patterns. Youtube link to watch the game. Download link for the full game (Windows Build).

Monogame C# Software Design Patterns Agile Development

Project explores the use of WebGPU, a new graphics API, for direct volume visualization of tetrahedral meshes.

This project began by the need to efficiently render complex, unstructured datasets using WebGPU’s feature set to implement direct volume visualization for a tetrahedral mesh. The renderer traverses the tetrahedral mesh sequentially, accumulating color and alpha values. This continues until the ray exits the mesh.

Renderer is capable to visualize the turbulence around a golf ball, but currently faces performance limitations when handling large meshes due to CPU-intensive pre-processing steps. Future work will focus on moving these pre-processing steps to WebGPU compute shaders to improve performance and enable the visualization of larger, more complex datasets.

Tetrahedral Volume Rendering - Research Report

Typescript WebGPU API WebGPU Shading Language HTML

Divides a given mesh into n number of subdivisions using Catmull-clark subdivision which smooths out low-poly meshes. Demonstrated expertise in computational geometry and mesh processing techniques, contributing to the refinement of low-poly models for improved visual fidelity.

Python Geometric Modeling

Executable 2D minigame space invaders minigame where you control a laser canon and try to shoot down incoming aliens! Follow the instructions on the repo to play the game!

FullStack .NET Dev

About Me

Experience

Projects

Education

Contact