Increasing student engagement through GenAI stories and quizzes
Choice Texts Case Study — AI Story Generation, LLM Prompt Engineering, UX Research with Students, and Personalized Reading Experience Design by Thom van der Doef
A generative-AI reading experience where students of all ages co-create their own story before taking comprehension questions. I led Product Design, Discovery and Delivery.
Company • eSpark Learning
Role • Lead Product Designer
Duration • 7 Months concept to launch
Team • PM, Product Designer, 2 Devs, 2 Content
Problem: Low student engagement
After COVID, eSpark saw a drop in student engagement, putting learning outcomes at risk.
Can we use generative AI to make our activities more meaningful for students, leading to more engagement and better learning outcomes?
What did we build?
To maximize impact across our curriculum, we started by transforming our Reading Comprehension activities—the highest-volume area and the biggest source of disengagement.
Our solution was simple but powerful: let students co-create the illustrated stories they would later read and answer questions about. This instantly boosted ownership, motivation, and attention going into the comprehension quiz.
That raised the key product questions that shaped our approach: Which choices are most meaningful for K–5 students? And how do we keep the activity focused on Reading Comprehension rather than story creation?
UX Process
- What experience would be most meaningful and engaging?
- What was technically feasible?
- Which subject or domain to focus on first?
- Led internal workshops to explore concepts
- Tested three ideas with students at CPS Melody and MAP
- Ran OpenAI experiments and vetted with Learning Design
- Clear, scoped concept: a standards-aligned story builder with comprehension questions
- How should the story-building flow work?
- Which small set of choices is most compelling?
- How do we guide students step-by-step?
- How do we build buy-in from the team?
- Tested a basic prototype with kids (CPS + afterschool program)
- Led a cross-functional synthesis workshop
- Summarized all research into recommendations + roadmap ideas
- Main concept & UX validated — kids loved it
- Clear direction on design and choice model
- Early skeptics became advocates after seeing real student reactions
- How does the end-to-end student workflow function?
- How to support readers and emerging readers?
- How to scale to other subjects and domains?
- How to generate millions of clean, school-appropriate images cost-effectively?
- Mapped the workflow in FigJam with LD, dev, and design
- Designed developmentally appropriate UI/UX
- Built a flexible conversational UI framework
- Engineered an image-generation pipeline using OpenAI + Leonardo.ai
- 9M+ unique illustrated stories created by students ages 5–15
- Highest-rated activity type in eSpark history
- Framework scaled to Informational Text, Decodables, Writing, and Math
Designing student choice
We wanted kids to feel like they have some ownership over their story and give them lots of choices, but the focus of the activity should remain on the reading and comprehension questions.
What choices are most meaningful for kids? How can kids create a meaningful story quickly? To find out, we conducted a series of design and testing sessions with 4th and 5th graders in Chicago Public Schools.
Limit story choices
We found out that these 4 choices gave kids the most sense of control and ownership: Main Character, Name, Traits and Story Setting.
Presets and open text input
This activity was for kids ages 5-13 with a wide range of abilities, so we needed a pattern that allowed non-readers and proficient readers both to make choices appropriate for their development
Older students can type a character description or setting.
Creating a flexible and scaleable system
Even though this activity was limited in scope (create a story, answer questions about the story), we expected to create other activities for Reading, Math and perhaps Writing, that would require many types of user inputs, system prompts and artifacts.
Conversational UI
To guide kids through the story-creation process, we opted for a conversational UI that allowed for a variety of age-appropriate user inputs.
Focus
This conversational UI can be shown or hidden to allow for focus on a playful back and forth with a character or the text artifact.
Other applications
Using the patterns and systems established in Choice Texts, eSpark was able to launch Choice Math, Themed Choice texts, Decodable Readers, eSpark Writing, and Genre-based Choice Texts in the same year. It has transformed eSpark’s curriculum offering with new content and subjects
Side-by-side view
We improved on the quizzing interface by having the conversational UI side by side with the story. This interface will allow for students to interact directly with the text as they respond to reading and/or writing prompts.
Generating Millions of Safe, Kid-Friendly Images
Generating one great image is easy. Generating millions of them for classrooms—safe, age-appropriate, unbiased, non-creepy, and affordable—is not.
Early prototypes that fed stories directly into image models produced inconsistent characters, strange artifacts, unpredictable safety issues, and rising costs as volume grew.
I solved this by designing a two-step pipeline. First, OpenAI produces a precisely structured, safety-constrained image prompt based on the student’s story. Then that cleaned prompt is sent to Leonardo.ai, which renders a consistent, kid-friendly illustration. This reduced artifacts, stabilized style, lowered per-image cost, and enabled Choice Texts to scale to millions of reliable, classroom-safe images.
Read more about the process here: Article: eSpark Learning Brings Personalized Learning to Life Through Leonardo’s Image Generation API
Translating student engagement into outcomes
9 Million+
Stories created
97%
student rating
+23%
NWEA scores
So far 1 million students created over 9 million unique, illustrated stories.
Choice Texts are the highest rated activity type in eSpark’s Curriculum
An internal study showed that students with a high exposure to Choice Texts scored 23% higher on 3rd party assessments.
“Seeing my most reluctant readers engaged [...] reading and talking about their stories, is fantastic! ”