At ElleHacks 2020, my team created an approachable iOS companion app that would teach young adults to recognize their emotions using natural language processing and provide resources to develop emotional intelligence.

Business Model

36 Hours



Many existing therapy apps on the iOS app store have limited input options for the user, are poorly rated, and don't provide resources to alleviate loneliness. We noticed, that while the audience of the therapy apps faced restrictions in expressing themselves, they were inquisitive and continued to track their emotions.

Our goal is to give students and young adults an experience that was simple, reduced input friction, and provided resources to navigate different emotions.

My Work

I conducted market research and competitive analysis. Using the pain points and constraints I focused on the flow of conversation between the app and users. I also prototyped the final product, and developed a scalabe business plan. 

Design Goals


Track Emotions

Understanding emotions helps us manage them. By being aware, we can make informed decisions, prevent or avoid triggers of negative emotions, and communicate clearly to get support.


Provide Resources

Learning how to manage emotions can have positive outcomes despite negative emotions. Resources can offer: reflection, perspective, acceptance, and help in making rational choices.


Reduce Input Limitations

Understanding emotions helps us manage them. By being aware, we can make informed decisions, prevent or avoid triggers of negative emotions, and communicate clearly to get support.


Identify Patterns

Learning how to manage emotions can have positive outcomes despite negative emotions. Resources can offer: reflection, perspective, acceptance, and help in making rational choices.


What do people think about loneliness?

Starting post-secondary means leaving behind valuable support systems. We wanted to understand how students dealt with this.

Target Users

We were most interested in people between the ages of 18-34. Students are the most vulnerable to loneliness.


We relied on two international students in the group for insights. We validated assumptions through secondary research.



Nearly 70% of students felt lonely over the past year.


Culprits include losing childhood friends & seeing picture-perfect social media posts.


40% of Canadians say they wish they had someone to talk to, but do not.


A change of environment can help combat feelings of isolation.


What other solutions already exist?

We need to understand which apps users are currently gravitating towards. What benefits do these applications provide our users? What are the disadvantages?

Comparative Analysis Chart

Self-help, therapy, and meditation apps are aimed to help people with their mental health.

Key Points


Young adults are looking to vent and release emotions without relying on pre-existing support systems.


We want to focus on companionship rather than therapy to help alleviate loneliness.


Mental health apps provide the tools for mental success but do not address and relieve emotions being experienced in the moment.


Many apps restrict input options. Users feel the system is robotic, limited, and cold.


What makes WoeBot an award-winning app?

Woebot strives to coach users through chats and "offers insights and skills to help them grow into their best self". Woebot aims to help users with everyday stress and anxiety & loneliness.


  1. Daily check-ins with pre-filled answers.
  2. Uses Cognitive behavioral therapy (CBT) to help users identify feelings.
  3. Links to external quizzes and videos to lead users to discover their thoughts.
  4. Remembers user’s history of past feelings and conversations.


  1. Asks too many questions.
  2. Restricted responses feel robotic and limited.
  3. Conversation flow is not natural, users may feel disconnect.
  4. Only focused on therapeutic solutions.


Developing a Persona and Early Sketches


Eden is a 22 year old moving away from his family to San Francisco to start his full-time career.


Final Application Design


1. Personalized Question

Prompt is focused on the user. UX writing similar to how friends and family would encourage and stimulate conversation.

2. Audio Input

Talking removes barriers, allowing the user to share raw feelings. Conversation starts with single button and feels natural.

3. History

Creates "journal-like" experience for users to track, revisit, and learn from their emotional history.

4. Recommendations

Insightful recommendations are accessible regardless of user's daily input. They will help the user learn about what to do in various emotional states.



Screening user's likes and dislikes will allow the algorithm to create recommendations based on the user's preferences.


Resources for the user to utilize on based on the emotions identified in the session, but are also accessible in the main navigation.



For accessibility reasons, Ami offers text input. Users can text with Ami if they require or prefer chatting instead of using the voice function.


Identifying Emotions

Through ML (sentiment analysis), Ami identifies that Eden is excited and joyful and celebrates with him.

Answering the "Why?"

  • Important when referring to history for recommendations
  • Helps recognize emotional patterns
  • Over time, creates self-awareness
  • Equipped with tools to develop emotional intellegence

Tracking Emotions

Ami further helps Eden with identifying the emotions its responding to Eden with.

Why is this helpful?

  • Idea of "talking to yourself like you're talking to a friend"
  • Help understand despite experiencing negative emotions, they must be kind to themselves

Users are able to talk, chat, vent, track, acquire resources, and learn from Ami.



Designing A Brand Identity Kit


My team resonated the feel with warm earth tones to reflect the connection to self. We chose shades of green in the colors, type, logo and imagery to represent self-growth.


Invisioning Ami's growth potential


Companies such as Spotify and Headspace can be embedded to suggest recommendations for their products and services that are catered towards our users’ needs. Other revenue streams can include affiliate links, streaming services, destinations (museums/cafe's). By using this model we can easily add more sponsorships from a diverse range of services.


What could Ami look like in the future?


Ami can be expanded into products through smart home appliances. Implementation in voice assistant devices, such as Google Home or Amazon Echo, Ami can deliver an impactful flow of conversation.


Lessons Learned and What I can Improve


Ami was a thrilling experience and I'm really glad we were able to complete research, designs, and prototypes within 36 hours. To validate our solution, I would conduct surveys, user interviews, and spend more resources to gain insights on our target audience to validate our problem space. I would facilitate usability tests between each prototyping stage to test the ideas.

View More Projects

WSIBUser Research, UI Design, Analytics

PlantellaUser Interface Design

PawshUser Research

Milton Public LibraryUser Research

©2023 P. Mehta
Designed with TLC ✨