Background Image

@thebigsleepjoe

Programmer • Web Developer • IT Tech

Lua
Python
JS/TS
Java
PowerShell
Media
Love2D
Modding
AI
ML
LLMs
Web Scraping
Data Analysis
Next.js
React
Tailwind
Bulma
CSS
HTML
LWJGL
Video Editing
Image Editing
Git
GNU/Linux
Windows
Networking
Scroll down

Hello!

Welcome to my personal website. This is where I host my portfolio, contact info, and blog.

This is largely my programming-focused showcase/portfolio. If you're looking for my resume, you canemail me with info (like who you are) and I may send you a proper CV.

I also have a blog! I don't write much, but if you like tech stuff, you can read ithere.


Programming Portfolio

TTT Bots

Lua
Modding

A bot addon for the popular Garry's Mod gamemode Trouble in Terrorist Town (TTT), and its derivative, TTT2. Check it out onGitHub.

Necesse Ambience

Java
LWJGL
Modding
Video Editing

A mod that brings custom-picked sounds into the game world of Necesse. This adds footsteps, animal sounds, and global ambience which is dynamically muffled when you move indoors.GitHub.

bigsleepjoe.com

Typescript
React
Next.js
Tailwind
GNU/Linux

The website you're currently looking at! This is built using Next.js, Tailwind, and elbow grease. As of writing, this is being hosted on my Raspberry PI 5 via Cloudflare tunnel.

The source code for this website can be found onGitHub.

YT Comment Analysis

Typescript
Python
AI
Machine Learning
Video Editing

I scraped tens of thousands of YouTube comments from over a thousand YouTube videos, sorted them by video category, and ran a few data processing algorithms on each comment.

I then compiled some interesting results and graphs into a video, which you canwatch here, or you canread the actual report here.

Pithee Scraper

Python
Web Scraping
AI
Data Analysis

I wrote a scraper for the websitepithee.com, which is a crowdsourced joke website by Jacksfilms. My scraper uses a login token to grab a bunch of random user posts and can collect large amounts of post data in a short amount of time. It can also collect winner posts at a much slower rate without needing a token.

My original intention was to use this data to train a classifier to measure if a joke was funny or not. Unfortunately, the winning/losing jokes were actually too similar for any classifier I tried to actually differentiate them (in a statistically meaningful way).