Personal Projects
A collection of software, research, and linguistics projects I've worked on. Where applicable, links to the associated GitHub repository are included.
Linguistics & NLP
-
Code to Analyze Intensifiers
A custom Python algorithm developed to collect and analyze intensifier data from transcripts of Spanish dialects. Used to support ongoing sociolinguistic research comparing intensification patterns across different Spanish-speaking communities. Has been used in multiple conference presentations and supports a paper published in the Proceedings of the Linguistic Society of America 11(1) (doi:10.3765/plsa.v11i1.6091).
-
An Online Dictionary of Louisiana Creole
A web-based dictionary for Louisiana Creole, a critically endangered language. The goal is to make lexical resources accessible online to support language documentation and revitalization efforts. Built in collaboration with members of the LSU Creole Club.
-
Spanish Future Tense Research
A Python-based project for analyzing expressions of the future tense in Spanish corpora. Includes scripts for data cleaning, grammatical tagging, and building structured dataframes from raw corpus data to support quantitative sociolinguistic analysis.
-
Praat Data Analysis Scripts
A collection of Praat scripts used for collecting and analyzing audio data in a study on dialectal variation and perception of English sounds. Handles batch WAV file processing and automates acoustic measurements that would otherwise need to be done manually.
-
A Short Spanish Podcast about Cultural Representation
0:00 0:00
Software & App Development
-
Vivan — Language Learning Platform
Originally built for learning Louisiana Creole (Kouri-Vini), Vivan has grown into a full language-learning platform where users can take predefined courses or build and share their own. Built with a small team as a CSC 4330 project at LSU, with course creation, collaboration, and 2FA support.
-
Multilingual Wordle
A web-based Wordle game, live at multilingual-wordle.fly.dev, supporting 70+ languages from Wiktionary — including Spanish, Russian, Arabic, Thai, and more. Handles grapheme clusters, abugidas, and RTL scripts correctly, with word lists cached locally after first download.
-
Swadesh Word-A-Day
A Python tool that scrapes Wiktionary's Swadesh list appendices across 200+ languages to build a vocabulary dataset, then sends a daily word-of-the-day push notification via ntfy.sh. Scraping is parallelized across 8 workers, supports filtering to the 30 most-spoken languages, and is designed to run on a cron schedule.
-
Natural Syntax LSP
Parts-of-speech, semantic-embedding, or dependency-parse highlighting for VS Code via a local Go LSP server running ONNX inference (BERT/MobileBERT, MPNet/MiniLM, or a biaffine UD parser). Hover any word for its tag, head/dependents, and a Wiktionary definition.
-
GuideOverride
Collaboration with Baton Rouge Community College to streamline student course-override requests using an LLM recommendation workflow. Featured in the Business Report LSU AI Showcase.
-
Out of the Abyss DM Companion
A local web app for running the Out of the Abyss D&D 5e campaign. Tracks NPCs, monsters, sessions, maps, and initiative, with a local-LLM chat assistant and semantic lore search powered by Ollama.