CorpusKey

CorpusKey

Scope

UX Writing

Information Architecture
Testing

Duration

A semester

Worked 3x a Week

Tools

Figma

Excel

Notion

Project Overview

Project Team: 2 UX Graduate Students and 2 UX Undergraduate Students

Background: CorpusKey is an AI-assisted course material creation tool that instructors can use to create textbooks, lesson plans, readings, and ppts. Users can input learning objectives and key points of what the academic materials will contain and the integrated learning language model (LLM) will generate content.

Goal: The goal of this project is to improve the user flow of generating content with AI and the language that confuses users. Most of the work will focus on the “Outline Management” page where the user creates an outline with the topics they hope to generate readings from.




Problem Space: CorpusKey’s current website is still in development and designed from the perspectives of the software developers. The website uses language and a complicated flow that is not easily understandable for users. 



Final Deliverables:

  1. Design of the Outline Management Page

  2. The project documentation

My Role:

UX Writing, Usability Testing, Heuristic Evaluation, Sketching, Wireframing, Prototyping

Milestone 1:
Understanding of current workflows of instructors

Goal: To fully understand the pain points that our new platform needs to address, we must first research how university professors currently generate content. 


Approach: We performed a comparative analysis to empathize and understand the features users look for and what already exists in course-creating platforms. Through research, we gained a better understanding of the field and then conducted interviews on the existing website.

Comparative Analysis

Interviews

Comparative Analysis

Milestone 1:
Understanding of current workflows of instructors

Goal: My team wanted to better understand what solutions already existed for providing accessibility in the makeup industry. We wanted to research both online shopping and makeup products themselves.

Approach: We performed competitive analysis of accessible apps in the makeup industry to find out solutions that exist already. We knew that Lauder had their own app called the Estée Lauder Virtual Make-up Assistant so we performed a heuristic analysis of it to find their strengths and weaknesses.

Comparative Analysis

Comparative Analysis

Comparative Analysis

Milestone 3:
Recommendations to improve information architecture

Milestone 3:
Recommendations to improve information architecture

Goal: Implementing a new information hierarchy for CorpusKey, particularly within Outline Management, is a major part of our project and one of the biggest issues with the current interface.


Approach: Test and evaluate a new wording system and information architecture that reduces the learning curve for new users.

Evaluation

Testing

Recommendations

App Features

Wireframe

Project Team: Anika Gupta, Arushi Gupta, Tiffany Lau


Background: Lauder’s ELC Accessible Beauty Virtual Hackathon wants teams to ideate/develop inclusive technology solution that enhances accessibility in the beauty industry.


Goal: Develop a functional application for an inclusive technology solution to help tackle accessibility challenges in the beauty industry. Solutions can address online or in-person shopping experiences, accessibility in UX design, equal access to beauty products, and more.





Project Overview

Problem Statement: In the beauty industry, there is a lack of accurate & user-friendly resources to help visually impaired people use and apply makeup. This makes it difficult for them to apply makeup and they have to seek alternative ways without help.


Final Deliverables:Utilize any relevant developer tool and build a project Include a link to functional code repositoryInclude a demo video (around 5 minutes)


Contribution Overview: Competitive Analysis, UX Research, Affinity Diagramming, Ideation, Wireframing, Prototyping

Milestone 2:
Evaluation of the current site interface

Goal: To ensure our site addresses and corrects the pain points faced on the site, we must analyze the issues that users might run into with CorpusKey’s current interface


Approach: We conducted a heuristic evaluation and usability testing to find the issues in the previous design in order to come up with design recommendations.

Heuristic Evalution

Usability Testing

Jump to F nal Design

i

Design Process

Menu

Menu

Milestone 1:

Understanding of current workflows of instructoes

Milestone 2: Evaluation of the current site

Milestone 4:

New UI Design

Milestone 3:

Recommendations to improve information architecture

Empathize

How do instructors create course material and what are their needs?


How do other online platforms handle AI generated course content?

Interviews:


Interviewed 2 Professors: Dr.B and Dr. H

about pain points and preferences

regarding tools and systems


Comparative Analysis:


Looked into 5 different online course-creation platforms



hover to see insights

Main Paint Points:

  • Course creation method: they modify existing material every semester.



  • Wary that AI based tools take away the creativity from the course creation process.

  • They struggle to understand how Brightspace works.


Preferences:

  • They like to prepare slide decks for their classes.



  • A tool would be helpful if it translated the intention of the instructor.

  • One of them suggested that the AI could help suggest learning objectives and paragraph headings to generate the outline.




Overall Findings:

  • Coursebox, Magic School, and Taskade offer AI-powered features to streamline course creation and management, but have some limitations in terms of content uploads, editing, and automation.

  • Canvas has a comprehensive course management but lack AI features, requiring more manual platform management.

  • Perusall stands out with its AI-based grading functionality, enabling collaborative text annotation and reducing the burden of manual grading for professors.

UX Writing:


Through the analysis we also focused on exploring the different UX writing within the platforms and how they structured their content outlines. Common words that were used included: sub-headings, chapter, topic, sub-topic and others.


Original Website Testing

What does the userflow of the website look like and what are the important tasks to include in testing?


What pain points do users encounter when using the current CorpusKey website?

Heuristic Evaluation:

We made annotations of issues and suggestions to incorporate in our design.




Cognitive Walkthrough:

I went through the website and marked up the actions in order to narrow down the tasks for usability testing.




Usability Testing:


Recruited 6 Purdue students and a teaching assistant to participate in testing which had a list of tasks for the participant to complete.



The tasks:




Affinity Diagram of the results:

The hearts and thumbs down were the sponsor's responses to the proposed solutions.

deleting project

creating the project

uploading files

editing the outline


Ideation & Prototype 1

Testing Prototype 1

How to improve the outline management page to be more intuitive?

Design Recommendations:


Through the usability testing and heuristic evaluation, we gathered all the issues and made a condensed list of recommendations to include in the prototype.




  1. New Terminologies to Heading: Heading 1, Heading 2, and Heading 3 inspired by comparative analysis


  2. Dropdown Preview: Our comparative analysis inspired a feature that allows users to see how a heading will appear in the document.


  3. Draggable Outline Element: The draggable outline elements allow users to easily adjust the position and orientation of the outline


  4. Progress Bar: We suggest incorporating a progress bar as it helps users track their current progress and understand which stage they are in.


  5. Nested Outline: The nested outline features hierarchical headings, enabling users to identify the different levels of hierarchy clearly.

Site New Update Evaluation!

Takeaway: After looking at the results of testing, it was clear that our progress bar design wasn't efficient and didn't communicate the information architecture clearly so it was time for a new design.

While the design team was unaware, CorpusKey’s dev team took in a few of our design recommendations and implemented them into the website like the progress bar.



Our design team wanted to receive feedback on the progress bar design.


Therefore, we conducted a second round of heuristic evaluation and a task analysis to identify new pain points users may be having with the site. Here are the main insights:

Milestone 4:
New UI Design

Goal: After learning from our previous design, we wanted to create a new information architecture for the outline management page that is more user-friendly.


Approach: We started to prototype a new solution by brainstorming and doing a showdown for us to get many options for features to add in the final design.

Brainstorming

Mid-Fi Showdown

Prototyping

Testing

Ideation & Prototype 2

What is a better prototype that encompasses all our feedback?


Brainstorming:

We decided to run a brainstorming session to think of more ideas.


Mid-Fidelity Showdown:

After coming up with the general structure of the page, we did a Mid-Fi showdown where all the group members made their own interpretations of the design.



During our session, we came to the consensus of going back on one of our previous ideas, which is a one page outline management system.


Rationale: The user can easily go back the previous step without thinking there is a number order.

Collaboration:

As a group, we went through them all and talked about the features we should include and things we shouldn't.


We all agreed that we shouldn't include a number system since the steps can be done in any order.

my prototype for mid-fi showdown

all the mid-fis my team came up with

Mid-Fidelity Prototype

After putting all our ideas together, me and another teammate drafted the mid-fidelity prototype so we can do testing with it.

Usability Test: Card Sorting

Final Design

In order to ensure that the final prototype was easily understood and accessible, we conducted card sorting activities with our user group (5 Graduate TA’s in different academics).


We used card sorting, tasking participants with arranging labeled cards into both an outline hierarchy and a chronological sequence of steps.  














Conclusion:

  • The participants were able to arrange the cards in a way that represents the information architecture followed by the redesigned website. This validates the information architecture followed in the redesign. 

  • The participants also followed the chronological sequence of steps of creating an outline in the same structure as the website thereby affirming our design approach.

This is our final design which is an interative prototype of our high-fidelity design so you can click through it. Our stakeholder was very satisfied with the result and this design is now in development!

Contribution Overview:

I was involved in interviews, testing, and prototyping because I wanted to work on my testing skills. I did the wireframing and prototype for our primary solution to show how we wanted the outline managment page to flow. I enjoyed keeping the team talking and collaborating by often asking questions to the team; by the end of it we became really close friends.


What I learned:

This project was special because while my team and I worked on giving the company design recommendations, we had to keep adjusting and testing when the site would recieve new updates. It is important to communicate and test before the developers recieve design suggestions so that their time is not wasted.


It was a new experience where I learnt how to be agile and flexible with designing and I saw how important it is to stay open to ideas instead of being fixed on one design as after testing with a different one you may get different results to what you expect.








Thank you for visiting!

Come say hi!

Directory

© 2024 Anika Gupta

Designed by Anika Gupta