Designing an AI-supported academic writing workspace for college students with dyslexia.

College students with dyslexia often develop personalized workflows across writing, accessibility, and productivity tools. We explored how an integrated writing environment could reduce unnecessary cognitive and interaction burdens while preserving students' control over how they write and use AI.

Project Info
‍Role
UX Researcher and Designer
Team
Loan Pham,
Vandan Patel
Yengkee Chand
Methods
Contextual inquiry,
think-aloud testing
‍
Output
Interactive prototype
Project Type
Click to see the prototype ↴
On this page
The Problem
The Research
Key Insights
Define The Opportunity
Designing the Experience
Reflection
The Problem

Students shouldn't have to build their own accessibility ecosystem just to complete academic writing

Students with dyslexia were using combinations of writing, speech, read-aloud, proofreading, and accessibility tools to support different stages of their workflow. While these tools provided valuable support, switching between them introduced additional interaction and cognitive burdens.

The opportunity wasn't to automate writing for students. It was to reduce unnecessary friction while preserving the strategies and control that already worked for them.
The Research

What we wanted to understand

  1. How students currently approach academic writing
  2. Where cognitive load occurs
  3. Which tools and strategies students rely on
  4. Where switching between tools creates friction
  5. How students perceive AI assistance
01 — Planning creates cognitive load
Students had difficulty translating ideas into an organized structure before drafting.
02 — Accessibility is already a system
Students weren't relying on one accessibility feature. They assembled personalized combinations of tools and strategies.
03 — Voice and read-aloud support more than accessibility
Speech input supported idea generation and writing flow, while read-aloud helped students comprehend and proofread their work.
Key Insights

Students weren't looking for technology to write for them. They were building systems that helped them think.

Define The Opportunity

How might we bring students' existing writing and accessibility strategies into one flexible workspace without taking control away from them?

Reduce switching
Bring essential writing and accessibility functions into one environment.
Support multiple modes
Allow students to type, speak, listen, read, and edit according to their preferences.
Keep students in control
AI should suggest rather than silently modify academic work.
Make AI transparent
Students should understand what AI contributed and decide what happens next.
Designing the Experience

We designed the workspace around the major stages of academic writing: brainstorming, organizing, drafting, proofreading, and revision.

Reflection

This project changed how I think about accessibility.

Rather than designing a single “accessible” configuration, I learned that accessibility often means giving people control over the strategies, modalities, and levels of support they use.

It also changed how I approach AI-assisted experiences. The goal isn't simply to automate a task. In this context, the product needed to make AI useful while ensuring that students remained authors and decision-makers.

If I continued the project, I would involve students earlier in defining the boundaries of AI assistance—particularly asking them what the system should never do automatically.

‍

Dariel Tatiana Davis ©2026