In this game, students are falling from an airplane and must find a way to survive their fall. They can collect objects to make a parachute or create an improvised cushion by strapping into a row of seats and spinning to increase air resistance. Their parachute must have a canopy large enough to slow their fall, a frame sturdy enough to hold their weight, and cords long enough to fully deploy their canopy and thick enough to withstand deployment. Safe Landing tests students' knowledge of aerodynamics, drag, and terminal velocity, forcing them to think creatively in order to survive.

Codex - Blender - Meshy.ai - Javascript - Three.js

This simulation integrates Gemini, Google's AI model, with Articulate Storyline to create a dynamic learning experience. Users have free-form discussions with four different students, discover the root causes of their behaviors, and work with those students to find solutions. Each student is programmed with a distinct personality and backstory using Gemini's Generative Language API and comes to life through Vyond animation.

Storyline - Vyond - Google Gemini API - Javascript

In this piece of microlearning built in Articulate Rise filled with graphics I created in Adobe Illustrator, learners follow a woman, Justine, over the course of one busy day and see how she organizes her time. At the end, learners complete a quiz that has them apply the Eisenhower Matrix to Justine's schedule to optimize her time.

Rise - Adobe Illustrator

This is an assignment I co-created as a teaching assistant, and it served as the final project for the course Introduction to Technoculture at UT Dallas in the Fall of 2023. It requires students document their use of a large-language model as they iterate on prompts until the model produces an acceptable analysis essay covering the course's core concepts.

Chat GPT 4.0 - Blackboard LMS

An online professional development course built on the Canvas LMS and used by Celina ISD in Texas to better prepare teachers to integrate technology into their classroom practices.

Canvas - Canva

Scale AI

Developed PhD-level adversarial evaluation tasks that resist surface-level pattern matching by integrating subtle semantic distractors, contradictory logic, and nuanced concept masking while requiring LLMs to engage in abstract reasoning, synthesis, complex problem-solving, and conceptual mapping.

Engineered prompts that melded abstract critical and pedagogical theories and concrete learning scenarios with objectively verifiable, unique, and automatically gradable ground-truth final answers to ensure scalable, automated evaluation of AI reasoning across educational domains.

The University of Texas at Dallas

Incorporated students’ backgrounds, career and academic goals, and prior knowledge into instruction and curriculum design, integrating backward design and inclusive pedagogical principles to enhance learning outcomes.

Cultivated trusting relationships with diverse student populations through culturally responsive communication and active listening, resulting in a 25% increase in student participation rates.

Implemented an anonymous student feedback system, leading to curriculum adjustments and a perfect student course evaluation score.

Developed assignments that effectively incorporated artificial intelligence, large-language models, multimodal composition, and other new technologies and media to bridge the gap between students’ interests and skills and courses’ learning objectives.