Corpus Key

2025

AI Implemented Workflow

In this project my team aimed to improve CorpusKey’s AI grading tool by refining the user flow, reducing confusion, and updating the prototypes that were constructed during previous semesters, specifically to be able to handle grading various assignments. Our focus is on enhancing rubric accuracy and editing features.

Role

Product Designer

Timeline

Spring 2025

Team

8 Designers

Platform

Figma

Role

Product Designer

Timeline

Spring 2025

Team

8 Designers

Platform

Figma

My Contributions

Retrospective

Researched previous semester projects to understand where our project could build on past insights

Led Meetings

Communicated with stakeholders in person to showcase ideas/ methods for feedback and open discussions

Ideation

Sketched possible user flows to implement a grading rubric for different types of assignments

Figma Mockups

Utilized wireframes to mock up a mid-fidelity prototype for further understanding of the user flow

Retrospective

Researched previous semester projects to understand where our project could build on past insights

Led Meetings

Communicated with stakeholders in person to showcase ideas/ methods for feedback and open discussions

Ideation

Sketched possible user flows to implement a grading rubric for different types of assignments

Figma Mockups

Utilized wireframes to mock up a mid-fidelity prototype for further understanding of the user flow

The Problem

CorpusKey sought an improved grading system user flow and a more user-friendly interface. In order to improve these areas, we redesigned the grading support system to be more accessible and understandable for users regarding the process of file submissions and retrieval. For example, understanding the types of assignments that can be graded and how the grading style should vary from each assignment type. In addition to redesigning the grading support, we will also refine the grading support system with implementations such as assistance in rubric creation and validation..

The Solution

Our solution consisted of 3 different grading methods and how they would look on CorpusKey's platform. We create a prototype as concepts for what CorpusKey would do during or after the grading process. This prototype was then handed off to CorpusKey for future developement.

The user is prompted to select an assignment type. They can input their assignment and guidelines in the large text box. By doing this, the AI within the site can generate a rubric for them. Even with a generated rubric the user can edit the rubric freely. The user is also able to test their rubric with student assignments to see if there is any errors with the criteria.

For rubric testing, three random student responses are selected and graded three times to provide the professor with an accurate assessment of the rubric’s effectiveness. After testing, the user can go back and edit the rubric OR they can continue to grade the rest of their assignments.

After the user chooses to grade all, a thumbnail of the slide is shown, with the corresponding feedback. The user can edit the feedback by adding or removing content as needed. Details on how the grade was calculated are displayed under 'View details”.

The Process

Journey Mapping

Using Figma we created the persona of Professor Byrd to better understand our user group’s perspective. By the end of the mapping, we identified a task list, possible user emotions, pain points, and improvement opportunities that Professor Bryd would undertake.

Journey Mapping

Comparative Analysis

We examined six platforms that included a grading process with rubric creation or integration

We analyzed each platform by looking at each platform’s strengths, weaknesses, unique features, and overall usability & accessibility

Comparative Analysis

We examined six platforms that included a grading process with rubric creation or integration

We analyzed each platform by looking at each platform’s strengths, weaknesses, unique features, and overall usability & accessibility

Comparative Analysis

We examined six platforms that included a grading process with rubric creation or integration

We analyzed each platform by looking at each platform’s strengths, weaknesses, unique features, and overall usability & accessibility

Ideation

To begin brainstorming, we sketched possible solutions inspired by our research leading up to this point. Below is one of my sketches for how a user could utilize a rubric within the CorpusKey platform.

Ideation

Hi-Fidelity Prototype

All of these activities allowed us to come to our final solution which you can find at the top of this page!

Hi-Fidelity Prototype

All of these activities allowed us to come to our final solution which you can find at the top of this page!

Hi-Fidelity Prototype

All of these activities allowed us to come to our final solution which you can find at the top of this page!

Impact & Reflection

Collaborating with a team of 7 others on a project focusing on the integration of AI was a great learning experience for me, especially in the time of rapid AI adoption in product design. I took on roles such as a mentor, meeting lead, planning lead, as well as a fellow UX designer among my peers. Because this project was in partnership with a startup, it will have an impact on the future state of CorpusKey and how their products evolve.