Try it here!

Project Overview

Aura's leadership team wanted me to design a voice user-interface for their Alexa Skill idea. Aura allows users to improve and monitor their mental health by tracking their moods on a daily basis. 


UX Design,

Voice Design, Research


2 weeks

Stakeholder Analysis

Understanding the product

In order to get a comprehensive understanding of the product concept as well as establish design expectations, I interviewed key decision makers. Here's what I learned: 

Aura wants to help users tackle mental health issues by utilizing Amazon's Alexa. 
Competitive Advantage
By utilizing voice interactions, Aura makes it quicker and easier to track moods.
Target Audience
Target users are those suffering from chronic or major depression.
I was given two weeks to complete the design and hand it off to developers.


Scoping the problem

Despite so many Americans struggling with depression, mental health is still an often overlooked social issue. As more voice-assistants enter our homes, there might be an opportunity to utilize this new interface to improve our emotions.  


American adults suffer from Major Depression Disorder (MDD) (ADAA, 2016).


Americans with MDD do not receive treatment (ADAA, 2016).


Americans own an Amazon Alexa device

(NPR, 2018).


Alexa users continue using new “Skills” after the 1st week (VoiceLabs, 2017).

User Interviews

Empathizing with users

I interviewed Alexa owners to understand common pain points with voice-assistants. I put an emphasis on learning about their struggles with the new interface. 

User Problem: "I hate Alexa apps that make me talk too much."

My Solution: Aura can have quick voice interactions by limiting affordances. 

User Problem: "Some newer apps have such robotic and dull responses."

My Solution: Study natural conversations and integrate these responses into Aura.

User Problem: "I can never remember what apps I've installed on Alexa!"

My Solution: End all interactions with a reminder to check-in tomorrow.

User Problem: "I've been trying to get in the habit of journaling but just can't do it." 

My Solution: Provide an opportunity to quickly record daily moods in just seconds.


Brainstorming design features
Affinity Diagram
Affinity Diagram
Working from home :)
Working from home :)
Affinity Diagrams

After conducting interviews, I was left with a ton of information. To organize all this, I utilized an affinity diagram and grouped user feedback to find common themes and patterns. This also helped me prioritize design features.


To simplify the interaction, I decided to limit functionality to two options: 

Log Daily Mood

Allows users to quickly record their daily mood or emotion using just their voice. 

Emotional Insight

Allows users to gain insight into their average mood throughout the week/month. 

User Personas

Modeling user needs

To model some of the needs and frustrations of users uncovered from the affinity maps, I created user personas. This not only helped convey my research to the client but helped them empathize with their potential users.  

Jessie user.png

Generating Utterances

Testing early to track responses

The next problem that arose was that I needed to figure out a way to account for the wide range of emotions that people express and feel 


My solution was to create a small prototype to hear the various ways people expressed their moods. I classified emotions into 3 categories (Good, Okay, Sad) with the goal to expand on these as we iterated the product.


I went to my local coffee shop and invited patrons to participate in usability tests. Each volunteer was told to perform a specific task (log a mood or hear average mood) and I observed how they interacted with the prototype. These early tests were fascinating, as there was so much variability in how people expressed their emotions.

Testing Notes

Classifying Emotions

Card sorting to understand mental models

To ensure that I was classifying emotions into correct categories, I performed closed card sorts. 

Result of a closed card sorting session

Improving User Flow

Handling various use cases

Once I was confident with my classifications, I began testing again. This time, my focus was on improving the user flow and accounting for the various routes users might take. 

Aura User Flow.png


Explaining the impact

Due to rapid prototyping, testing and iterating, I was able to present the voice design to Aura's development team on time. I found that out of the 10 final usability tests, 100% of participants successfully logged their mood.


Our voice is the most natural communication tool that we have. With speech-recognition technology improving at such a growing rate, understanding how to design voice experiences is a great idea for UX designers. Utilizing voice commands can result in quicker, more intuitive and delightful experiences for users. Overall this was a great learning experience for me, and I welcome others to start designing voice interfaces as well!

User Interaction