Hi,
I'm Kezia and I am
a Linguist

About Me

I care deeply about language—how it moves, how it shapes us, and how it helps people navigate the systems they rely on. Whether I’m training AI models, designing educational tools, or building multilingual support systems, I bring curiosity, precision, and purpose to everything I create.

I’m a senior at UTA majoring in Linguistics with a minor in Spanish. As a nontraditional student with a professional background, I draw on years of hands-on experience in ERP rollouts, immersive learning simulations, and user support. I recently pitched an AI-powered language platform to the National Science Foundation, blending my interests in NLP and equity-focused design.

My work bridges backend logic and user-centered design. I prioritize heritage speakers, code-switchers, and multilingual users—people who live at the intersections of language and identity.

I design for connection, accessibility, and ease. Technology should fit into everyday life and make it better.

See My Bookshelf ↓

Projects

G-U-E-S-S (Association Game)

An association-based guessing game where users attempt to uncover the AI's secret word through conceptual links. While the interface and logic are still being refined, core functionality is in place. This project reflects my broader interest in building games that double as tools—for data collection, personalized training, and creating joyful learning experiences.

IPA Cup Toss

A gamified study tool that challenges users to identify the shared IPA (International Phonetic Alphabet) symbol between word pairs. Correct answers earn a chance to toss a virtual ball into a cup, blending phonetics practice with interactive fun. Designed for early Linguistics students and anyone looking to refresh their IPA knowledge. The card set continues to expand as the project evolves.

L’Oréal Smart Product Advisor

An AI-powered chatbot that offers personalized product advice across L’Oréal’s skincare, haircare, and makeup lines. Built using the OpenAI API, this assistant understands natural language, supports multilingual users, and links directly to L’Oréal’s official tools—all in a sleek, mobile-friendly interface aligned with the brand’s aesthetic.

Regex Romance Language Detector

A lightweight NLP pipeline that identifies spoken Romance languages using regex-based morphological clues (like ción, zione, ção, and more). Combines Whisper transcription and linguistic pattern-matching. I'm planning to add a Naive Bayes model for probabilistic classification soon!

Thirst Quest

A browser-based mini game developed in partnership with charity: water as part of a Summer 2025 internship. Players race against the clock to collect clean water drops while avoiding pollutants. Designed to raise awareness about global access to clean water through engaging, interactive gameplay.

NASA Space Explorer

A responsive web app that fetches data from NASA's Astronomy Picture of the Day (APOD) API. Users can explore stunning daily space images, learn their scientific context, and toggle between image and video media types. Built with HTML, CSS, JavaScript, and async API calls.