Projects
Venatio — Artemis II Mission Visualizer
There was no interactive, real-time visualizer for NASA's Artemis II mission that was accessible to the public in a browser. Venatio fills that gap by providing a live 3D visualization of the spacecraft's trajectory and mission phases, making the mission tangible for anyone curious enough to look it up.
Technical Stack: JavaScript, WebGL, HTML/CSS. Entirely front-end, deployed as a standalone web application.
Process & Reflection: This was my first large-scale project. I had to learn how to structure a complex JavaScript codebase without it collapsing under its own weight. The hardest challenge was getting the math right and rendering it smoothly in the browser. Debugging the mobile version of Venatio took lots of trial and error, but in the end I got a product that I am proud to share.
Link: venatio.app
EduFlash — AI Flashcard Generator
EduFlash is a community-powered library of educational resources where anyone with internet access can learn, share, and create lesson plans. Unlike traditional platforms, EduFlash empowers students, educators, and community members to upload their own guides, give feedback, and verify accuracy through ratings, discussions, and a trust factor system.
Technical Stack: Python, Flask, external AI/NLP APIs, HTML/CSS.
Process & Reflection: This was my first real exposure to web APIs and backend development with Flask. Going into the hackathon I had never used Flask before, so I had to learn routing, request handling, and API integration all under time pressure. The biggest challenge was structuring the API calls to reliably produce clean, usable flashcard output — the responses needed consistent formatting, which required careful prompt engineering and response parsing on the backend.
FlashBurn — USB Image Burner
The Unix dd command is powerful but
unforgiving — a single typo in the output device path can wipe the wrong disk with no
warning. FlashBurn wraps dd in a simple, safe CLI interface that makes
burning bootable USB drives faster and harder to mess up.
Technical Stack: C++
Process & Reflection: FlashBurn was written entirely without AI assistance, just me. That constraint made it one of my most educational projects. I had to actually understand every line rather than accept generated code on faith. The main challenge was correctly handling device enumeration and write verification across different Linux environments without relying on any third-party libraries.
Musi Playlist Downloader
Musi is a popular mobile music app with no official export or download feature. There was no existing Python tool to download Musi playlists, leaving users unable to back up or port their playlists anywhere. This project fills that gap, it is the only Musi playlist downloader.
Technical Stack: Python, BeautifulSoup, requests, yt-dlp.
Process & Reflection: The core challenge was reverse-engineering how Musi serves playlist data, there is no public API, so I had to learn BeautifulSoup and web scraping to extract track information from the page structure. Once I had the track list, I piped it into yt-dlp for the actual downloads. The trickiest part was handling edge cases: tracks with unusual characters in their titles, missing metadata, and rate limiting from the source.