I'm an academic (researcher) and developer focusing on Julia Lang packages, Electron.js apps :electron:, social physics simulations 🏙️, and useful tools. Below, I've organized my key repositories by category. Check them out—many include detailed docs, examples, and ways to contribute.
These tested tools have been peer-reviewed by the Journal of Open Source Software (JOSS). Great for data processing, NLP benchmarks, and more.
-
WunDeeDB.jl : A zero-config embedded vector database with WAL support. Useful for lightweight AI/ML embeddings. Ideal for vector search that scaled with transactions (which come with a cost for the confidence in data integrity) (JOSS-reviewed; in General registry)
Stars Forks
tip:using Pkg; Pkg.add("WunDeeDB")-try the examples for instant vector search. -
LMDiskANN.jl : Julia implementation of Low Memory Disk ANN for efficient nearest-neighbor searches. (JOSS-reviewed; in General registry)
Stars Forks- tip*:
using Pkg; Pkg.add("LMDiskANN")Ideal for large datasets and speed where transactions are not needed.
- tip*:
-
KeemenaPreprocessing.jl : Preprocessing for text data: cleaning, normalization, vectorization, tokenization and more. (in General registry)
tip:
using Pkg; Pkg.add("KeemenaPreprocessing"): Great for many text preprocessing tasks and also allows it to perform with controlled memory when processing large text corpora. -
BenchmarkDataNLP.jl : Generates synthetic text (e.g., via Context-Free Grammars) with tunable complexity to test NLP methods like LLMs.
Newcomer tip: Parameterize complexity to stress-test your models on less complex data before going into more complex natural corpora.
Two finished desktop apps for media tagging and organization. Built with Electron.js for cross-platform use—download releases for Linux/Windows.
-
Tagasaurus : Tag your planet with semantic search and machine learning. Search and annotate your media files with ease keeping all the data locally. Stars Forks
-
TagasaurusTagasaurusMemetic : The original "Tag your Planet" app—gateway to your semantic multiverse. Here memes and emotions take a first place position in search with specific features to allow for users to search for the impact of memes on certain keywords by utilizing a bi-partite graph. Stars Forks
Repos with code implementations from my research papers on social dynamics, entropy, and models like Schelling. Useful for simulations in physics/social sciences—many include Jupyter notebooks for reproducibility.
-
Eurovision : Investigating the collusion between countries for the complete historical data trajectory of the Eurovision Song Contest Stars Forks
featured in the economist magazine online and in print: https://www.economist.com/blogs/graphicdetail/2018/05/daily-chart-7 , https://www.economist.com/news/europe/21742117-continents-annual-singing-tournament-becoming-increasingly-partisan-neighbourly-voting . tip: Run the notebooks to visualize the one-way and 2-way (collusion) biases. -
EurovisionBias : Investigating the collusion between countries with both the positive and negative biases. So that the tendencies for countries to allocate significantly more or significantly less can be seen. Stars Forks
tip: Run the notebooks to visualize the full range of biases. -
Schelling Consistent : Addresses the issue of a decreasing entropy trajectory in the canonical Schelling simulation by introducing a dual dynamic with monetary exchanges so the arrow of time is respected. Stars Forks
-
Rat Utopia Modeling : Explores how the population collapse in the Rat Utopia experiment (Universe 25) can be simulated through a degeneration of the entropy of the hierarchy given the full visibility the rats had in the enclosure. Stars Forks
Handy packages and templates for everyday productivity, like LaTeX setups.
-
Nighttime LaTeX Template : A sonar themed dark-mode friendly LaTeX template for documents—reduces eye strain at night. Newcomer tip: Copy the .tex file and compile with pdflatex; perfect for late-night writing.
-
UCF Style LaTeX Slides Template : Beamer template in University of Central Florida style for presentations.
Newcomer tip: Customize the department logo.
- Star ⭐ or fork repos you're interested in!
- Open issues/PRs if you spot improvements.
Thanks for visiting! 👏