Objective

The goal of this phase is to combine the insights from the previous interviews and create a persona that represents all users to further evaluate their challenges, goals, and needs. During the define phase, we also decided whether our target audience really needs a solution to our problem. Some problems are better solved without integrating AI, so we also decided whether artificial intelligence is required in our proposed solution.

User Persona and Journey Maps

From the real-life interviews that we conducted, we were able to decipher what our persona, Garnet’s, needs, goals, and frustrations were. We wanted our persona to be familiar with scheduling tools, but not use them on an extreme level, because we found out that most of our interviewees only used their tools on a basic level through primary research.

This semester, we are being more critical about focusing on developing the problem, user needs, and AI and tech needs, so we utilized our resources and had ChatGPT as a teammate, using prompt engineering. For the user journey map, we wanted to emphasize that our target audience is college students, so the 6 stages were based on what a typical semester would look like. We asked ChatGPT to help us with the actions, thoughts, and a couple of design opportunities. With all of the answers that it provided us, we made sure to go back and verify the accuracy of the responses. My teammates and I decided the touchpoints, thoughts, and emotions based on the actions and goals of Garnet.

Evaluating User and AI Needs

Reflecting on the primary and secondary research, we gathered the findings and developed 2 key needs: Users need a way to not forget to eat or avoid sleep because of busy schedules, and users need a more efficient way to schedule social events with others.

With this worksheet, our team deciphered whether AI was really needed through voting on instances when AI would be better V.S. when it would not be. This method allowed us to discover different ways that AI could be useful that we haven’t thought about, and it required us to think more about the restrictions and challenges that come with AI, like how it can hallucinate and provide misinformation. This exercise helped us decide how AI would be integrated into our app.

After we decided how AI would be integrated, we evaluated the data requirements to train the AI with this worksheet. This process allowed us to think critically and realistically about what data the AI needs in order to be successfully integrated into our solution.

Challenges and Reflection

The biggest challenges of this phase lie in the evaluation of the user and AI needs portion. In the beginning, it was a little difficult to understand the vocabulary of the worksheets and what we needed to answer, but eventually, it became clearer with the help of ChatGPT. We were unsure about what “Key features needed in dataset” and “Key labels needed in dataset” meant, but with the answers that ChatGPT provided us, we gained a better understanding of what goes into those categories.