Pet Embedder
An AI-powered tool to get more pets adopted.
Role: UX design consultant for the Computational Outreach Research Lab at Virginia Tech
Developer and Lab Director: Ivan Hernanddez
Scope: 17 weeks
Tools: Figma and Qualtrics
Processes: User Research (Interviews & Surveys), Requirement Gathering, Wireframing, Lo-Fi and Hi-Fi Prototyping, Affinity Mapping, Visual Design, Usability Testing
Background
In his research, Dr. Hernandez found that the most significant factor for potential adopters when searching for a pet is morphology. In other words, how a pet looks is the driving factor behind their search for a pet.
Initial Concept
Pet Embedder is a tool powered by machine learning that streamlines the search for a pet in two steps:
Drag and drop a photo of a dog or cat that resembles the desired breed.
Scroll through the top morphological matches in pet shelters near the user. The closest matches will be shown at the top.
Wireframing, Requirements Gathering, and research review
At our first meeting, Dr. Hernandez gave a high-level overview of the tool, laying out the goals of the project, required features, and began rapid wireframing. l gathered requirements for the redesign and introduced design thinking to the team, emphasizing empathizing with users as the first step.
For this project, we used user interviews and surveys as our user research.
Initial interface before design
Before my onboarding, Dr. Hernandez had gathered data from potential pet adopters in the New River Valley region (Southwest Virginia).
After reviewing the data, we came to a consensus on the biggest priorities of potential pet adopters: proximity to their home and appearance.
The scope for this project, for now, is dog adoption.
More user research - expanding my understanding of the audience and landing on the correct problem(s) to solve
I established a new direction for user research: checking my assumptions, understanding user motivations, and uncovering their goals. I built and launched the survey using Qualtrics to 111 users, with unbiased, open-ended questions.
I discovered that 76% of people surveyed didn’t have a photo available to drag and drop of their ideal pet. This data informed my design decision to add an option in the user flow to select a photo from popular breeds.
