Home
About
Contact

Start Case Study View Prototype

Company: Reality AI Lab

Role: UI/UX Designer

Team: 5 Designers

Timeline: 3 day Sprint

NAVIGATE

Overview
Design System
Competitors
Solution
User Flows
Prototype
Outcome

Project Overview

In this 3 day competition, my team and I were tasked with crafting a next-generation AI Integrity Dashboard for the Marvel AI Platform. The dashboard needs to be intuitive & educator-friendly, while fostering trust, transparency, & fairness in AI detection.

Design Question

How might we design an AI-powered integrity dashboard that not only detects AI-generated content but also empowers educators with clear, explainable, and ethical decision-making tools?

Reality AI Lab Laptop Mockup

The Challenge

Provide explainability and transparency, showing why an essay is flagged as AI-generated
Offer visual insights that help teachers differentiate between fully AI-generated text, AI-assisted work, and authentic student writing
Maintain student trust by ensuring AI detection tools are fair, accurate, and non-disruptive to learning

Reality AI Lab Design Challenge

Design System

Competitor Analysis

Copyleaks Competitor Image

Copyleaks Integrity Dashboard

Identifies AI-generated content, including text produced by models like ChatGPT, Gemini, and Claude.
Provides comprehensive analysis explaining why text has been identified as AI-generated, enhancing transparency and understanding of AI patterns.
Detects various forms of plagiarism, including paraphrasing, image-based text plagiarism, and character manipulation.

Pangram Competitor Image

PanGram Integrity Dashboard

Utilizes a sliding window approach to individually assess segments of text, assigning each a specific AI likelihood score to pinpoint AI-generated sections.
Allows users to upload documents for analysis and maintains a chronological history of submissions, with options to filter, export, and share results.
Displays the highest confidence level that any part of the text is AI-generated, providing a clear measure of AI presence.

Turnitin Competitor Image

Turnitin Integrity Dashboard

Provides detailed reports that identify likely AI-written text, offering educators the information needed to determine appropriate actions.
Tailored to analyze academic writing, the system is adept at distinguishing between human and AI-generated content in student submissions.
Integrated within the Similarity Report, this feature highlights portions of text that may have been generated by AI tools, such as ChatGPT, providing a percentage estimate of AI-generated content.

Reality AI Solution

Reality AI Educator Dashboard Image

Educator Integrity Dashboard

Supports human verification, allowing educators to override AI decisions by leaving check boxes unmarked.
Provided color key to highlight the probability of plagiarism by percentage in specific lines of text.
Add comments function allows educators to communicate with students on specific assignment documents.
Sources section in the bottom right panel allows educators to see the plagiarism line and the specific web url linked to the text.
Full report button allows educators to see a full breakdown of AI text & human text percentages as well as plagiarism types identified.

Reality AI Student Dashboard Image

Student Integrity Dashboard

Total lines in document, lines of plagiarism, educator comments, & overall score for the assignment are displayed on student dashboards.
Flag icon allows students to view educator comments and reply to them for fast communication.
Clicking on highlighted text opens a modal and gives students suggestions for improving the text & adding citations.
Expanding the modal gives students a more in depth view for why the work work was flagged. This feature allows students to make suggested changes to assignments before or after turning it in.

Educator Account User Flow

Student Account User Flow

Educator Dashboard UI

Student Dashboard UI

Outcomes & Takaways

Competition Winners Screenshot

What I Achieved: I contributed to solving real-world challenges by implementing strategies that allow educators to check a student's academic integrity efficiently without disrupting the learning process. My team and I were successful in winning the design competition.
What I Learned: I learned how to organize, communicate, and execute a plan for building out a successful interface in a short amount of time.
Experience Gained: I gained extensive experience in working with a team of designers, following a design system, and executing a plan to create a successful interface based on user research and competitor analysis.