User flow for Prototyp
User flow for Prototyp
User flow for Prototyp

Project Status

Shortly after delivering this last update, I did not receive any updates regarding if this idea was going to become more than just a concept for proposal. I found out eventually that there were internal decisions made that resulted being edged out of involvement in this altogether. I was brought back in when there was a need because I was the most familiar with the idea and the person who made updates to the previous work that I had done was unavailable. These updates were to:

  • Adjustments user flows and UI elements that moved away from the intended use of Prototyp

  • Branding updates as this concept was being shown to another brand in the same field

Adobe Creative Cloud

One of the things I prioritized with this design was ensuring that users were able to jump into using this with minimal onboarding. To do so, I asked what applications were currently part of the users normal work flow. Adobe Acrobat was the answer. After looking at Acrobat, I took note of how things were structured to maintain order and also maintain quick accessibility to key tools.

Project Goal

For this project, a developer coded an AI algorithm to scan documents from a database for sentences called claims within a designated match percentage. The scan would be based on a new document enabling users to see if something said in that document was previously approved for use. While the AI algorithm was in MVP state, I was brought in to translate it into a UX framework and interface that would match the AI functionality.

How might we create a user-friendly interface for an AI model to significantly cut down users' manual document search time?

How might we design a user experience interface for an AI model that effectively reduces the time users spend manually searching through documents, enhancing efficiency and streamlining information retrieval processes?

Inspirations for design

The user base is composed of individuals who use Adobe Acrobat daily to complete their work. To reduce the amount of onboarding time required to learn a new program, I looked at several Adobe programs(Photoshop, Illustrator, Adobe Acrobat) for IA and structural insight when presenting information. I also took a look at Grammarly as Prototype worked similarly for identifying and highlighting new information.

Inspirations for design

The user base is composed of individuals who use Adobe Acrobat daily to complete their work. To reduce the amount of onboarding time required to learn a new program, I looked at several Adobe programs(Photoshop, Illustrator, Adobe Acrobat) for IA and structural insight when presenting information. I also took a look at Grammarly as Prototype worked similarly for identifying and highlighting new information.

Inspirations for design

The user base is composed of individuals who use Adobe Acrobat daily to complete their work. To reduce the amount of onboarding time required to learn a new program, I looked at several Adobe programs(Photoshop, Illustrator, Adobe Acrobat) for IA and structural insight when presenting information. I also took a look at Grammarly as Prototype worked similarly for identifying and highlighting new information.

Questions for the User

For this project, there was someone on the team who acted as the voice of the user. They provided perspective on current processes, goals for users, and the impact that we would have with this solution. Some questions that I had for them included:

  • What tools are currently used in the review process and for proofing documents for changes?

  • Would it be optimal to show multiple results for each claim?

  • What should the output for this be?

  • Does the AI algorithm scan the entire document or just what is different/not found in its database?

Questions for the User

For this project, there was someone on the team who acted as the voice of the user. They provided perspective on current processes, goals for users, and the impact that we would have with this solution. Some questions that I had for them included:

  • What tools are currently used in the review process and for proofing documents for changes?

  • Would it be optimal to show multiple results for each claim?

  • What should the output for this be?

  • Does the AI algorithm scan the entire document or just what is different/not found in its database?

Questions for the User

For this project, there was someone on the team who acted as the voice of the user. They provided perspective on current processes, goals for users, and the impact that we would have with this solution. Some questions that I had for them included:

  • What tools are currently used in the review process and for proofing documents for changes?

  • Would it be optimal to show multiple results for each claim?

  • What should the output for this be?

  • Does the AI algorithm scan the entire document or just what is different/not found in its database?

For this project, there was someone on the team who acted as the voice of the user. They provided perspective on current processes, goals for users, and the impact that we would have with this solution. Some questions that I had for them were:

  • What tools are currently used in the review process and for proofing documents for changes?

  • Would it be optimal to show multiple results for each claim?

  • What does the output for application look like?

  • Does the AI algorithm scan the entire document or just what is different/not found in its database?

The Problem Space

The Problem Space

Directions for functionality

  • Large main focus for content

  • Smaller half-size focus for document feedback and comments

  • Toolbar of common tools including a start process button

  • Chunk information by creating areas of focus for the document being reviewed.

  • Find a way to highlight comments/suggestions currently being reviewed

  • Include export options

Notes and initial ideas during brainstorm

Notes and initial ideas during brainstorm

Grammarly

During the initial conversations about this project, Grammarly stood out to me as a comparator for overall functionality. Prototyp as a whole had a different intended use however a lot of the basic functions and intended outputs overlapped with Grammarly. Specifically, the ability to scan a document and highlight specific information flagged by the system based on parameters set by the developer.

During the initial conversations about this project, Grammarly stood out to me as a comparator for overall functionality. Prototyp as a whole had a different intended use however a lot of the basic functions and intended outputs overlapped with Grammarly. Specifically, the ability to scan a document and highlight specific information flagged by the system based on parameters set by the developer.

Wireframes

Wireframes v2

After presenting a working Figma prototype demonstration to our client, they provided us with insight into ways to improve upon our idea. This included:

  • Pivot from an application that would allow editing within the Prototype to a companion program

  • Simplify the concept to only scan new documents and output matches from the internal database

Client presentation

Client presentation

Sample sketches

Project Overview

Prototyp is an AI-based application that scans information from uploaded documents and matches them to older documents in a digital database eliminating the need for a manual search. The purpose was to allow easier proofing and cross-checking if the information was previously approved by a company for consumer-facing media such as brochures and other information-based resources.

Please note that all mentions to and of clients by name have been removed or obscured to adhere to NDA restrictions.

Role: Product Designer

Tools: Figma, Miro

Timeline: 5 weeks