Healthily
Healthily
Optimising healthcare services through AI
Winning a 72-hour in-person AI hackathon with a redesign of the symptom checking experience for users, with an empathy-first approach.
■ Overview
Duration
72 hours
Tools
Figma, Dribbble, Mobbin, Loom
My role
Team Lead, UX/UI Designer, Branding
Team
Sole UX/UI designer
1 Business Strategist
Duration
72 hours
Tools
Figma, Dribbble, Mobbin, Loom
My role
Team Lead, UX/UI Designer, Branding
Team
Sole UX/UI designer
1 Business Strategist
Design Challenge
Healthily is a mobile app that is currently in the process of releasing a second product, an AI Symptom Checker.
Working closely with doctors, medical experts and software engineers, Healthily has developed a Symptom Checker algorithm that is 93% accurate in identifying potential health issues and recommending next steps. This AI is designed to redirect inaccurate streams of users within the patient and healthcare pipeline, based on their symptoms and the services they need.
However, despite its data-driven accuracy, the interface and design of the product has users weary and frustrated. Our thorough evaluation of the problem space, coupled with a design-driven, no-code solution created in just 72 hours, secured us the victory.
Presenting our winning solution
The problem
We know that users and healthcare providers are frustrated. A total of 60% of insured people are unclear of which services they should use:
30% of patients using telemedicine services are doing so when they can be treated with self-care;
30% of patients should have sought a face-to-face appointment instead.
In coming up with a problem to this solution, any LLM use cannot triage or provide guidance on treatment itself.
The solutions
Lily is a personal Health Detective. With Lily, users can check their symptoms and be informed of potential causes and advised on next steps.
As a “detective,” Lily is an AI that does not diagnose, rather works alongside the user as an empathic figure to get to their bottom of their symptoms and understand their next steps.
Duration
72 hours
Tools
Figma, Dribbble, Mobbin, Loom
My role
Team Lead, UX/UI Designer, Branding
Team
Sole UX/UI designer
1 Business strategist
Design Challenge
Healthily is a mobile app that is currently in the process of releasing a second product, an AI Symptom Checker.
Working closely with doctors, medical experts and software engineers, Healthily has developed a Symptom Checker algorithm that is 93% accurate in identifying potential health issues and recommending next steps. This AI is designed to redirect inaccurate streams of users within the patient and healthcare pipeline, based on their symptoms and the services they need.
However, despite its data-driven accuracy, the interface and design of the product has users weary and frustrated. Our thorough evaluation of the problem space, coupled with a design-driven, no-code solution created in just 72 hours, secured us the victory.
Presenting our winning solution
Design Challenge
Healthily is a mobile app that is currently in the process of releasing a second product, an AI Symptom Checker.
Working closely with doctors, medical experts and software engineers, Healthily has developed a Symptom Checker algorithm that is 93% accurate in identifying potential health issues and recommending next steps. This AI is designed to redirect inaccurate streams of users within the patient and healthcare pipeline, based on their symptoms and the services they need.
However, despite its data-driven accuracy, the interface and design of the product has users weary and frustrated. Our thorough evaluation of the problem space, coupled with a design-driven, no-code solution created in just 72 hours, secured us the victory.
Presenting our winning solution
The problem
We know that users and healthcare providers are frustrated. A total of 60% of insured people are unclear of which services they should use:
30% of patients using telemedicine services are doing so when they can be treated with self-care;
30% of patients should have sought a face-to-face appointment instead.
In coming up with a problem to this solution, any LLM use cannot triage or provide guidance on treatment itself.
The solutions
Lily is a personal Health Detective. With Lily, users can check their symptoms and be informed of potential causes and advised on next steps.
As a “detective,” Lily is an AI that does not diagnose, rather works alongside the user as an empathic figure to get to their bottom of their symptoms and understand their next steps.
The problem
We know that users and healthcare providers are frustrated. A total of 60% of insured people are unclear of which services they should use:
30% of patients using telemedicine services are doing so when they can be treated with self-care;
30% of patients should have sought a face-to-face appointment instead.
In coming up with a problem to this solution, any LLM use cannot triage or provide guidance on treatment itself.
The solutions
Lily is a personal Health Detective. With Lily, users can check their symptoms and be informed of potential causes and advised on next steps.
As a “detective,” Lily is an AI that does not diagnose, rather works alongside the user as an empathic figure to get to their bottom of their symptoms and understand their next steps.
Constraints
Financial
Financial
Financial
No budget to invest in marketing campaigns
No budget to invest in marketing campaigns
No budget to invest in marketing campaigns
Time
Time
Time
Limitations on in-depth user research, user testing
Limitations on in-depth user research, user testing
Limitations on in-depth user research, user testing
Technological
Technological
Technological
No software engineering or data team to develop a code-based solution
No software engineering or data team to develop a code-based solution
No software engineering or data team to develop a code-based solution
■ Solutions breakdown
PRODUCT MARKETING WEBSITE
PRODUCT MARKETING WEBSITE
Discover
Discover
Clear and friendly copy that educates the user on the benefits of the AI in an approachable and welcoming way
Therapeutic colours and rounded shapes make distressed users feel at ease
Social proof (e.g. quotes, mentions in publications and app numbers, etc.) provide a people-centred focus and build trust in the product
ONBOARDING
ONBOARDING
Learn
Learn
Adding a brief and informative onboarding process to ensure distressed users have clarity around product usage and expectations
Imagery with eye contact and copy in the third person emphasise human connection
Tiles with annotated images provide quick and easy to digest instructions on product usage
ASSESSMENT AND REPORT
ASSESSMENT AND REPORT
Check-up
Check-up
Split screen divides text from dynamic assets that update with user inputs, providing visual confirmation of AI processing information to reassure the user
Helpful icons provide context to the user about the stage and nature of the assessment, as well as the remaining time to completion
Informational tooltips provide users with additional explanation regarding specific aspects of the assessment process and build reassurance
Skeuomorphic design of the final report gives authority and builds trust in the quality of the assessment
ALIGNING ON THE PATH FORWARD
Building a healthy foundation
We started by identifying an accountability buddy at the hackathon, who took on an informal role of Product Manager. We set out a timeframe for ourselves to make sure we covered the necessary bases in the time available. We tasked our PM to keep us on track, ask us critical questions, as well as be a bouncing board for feedback.
■ Design process
SPOTLIGHT ONE
Making the user feel heard
What are ways in which we can make the user feel heard, when inputting data? How can machine learning validate how users feel? We explored ways in which we can design for users seeking "reassurance."
EXPERIENCE MAPPING
Pluse check
With an understanding of user expectations when it comes to healthcare services, and certain assumptions validated, we used this context to conduct product research to situate ourselves in their minds. This gave us perspective when evaluating the current symptom checking process that Healthily offers — currently marketed by the company as The Symptom Checker. We evaluated its features, functionalities, user interface, and user experience and identified areas of opportunity and that require further exploration.
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
How might we support patients to access the correct healthcare channels?
PROTO-PERSONA
Angelita's journey
A Healthily user who self-identifies as a hypochondriac
Often worried and anxious about her health
Seeks reassurance from friends, family and healthcare professionals
Looking for quick answers to her health concerns
Has been misdirected by healthcare providers in the past
Healing the open wound
SECONDARY RESEARCH
Expanding on empathy
The integration of AI should not compromise essential human qualities like empathy that patients seek in their interactions with healthcare professionals.
Currently, the product lacks elements of understanding and compassion, not only throughout the symptom checking process, but also in its overall look and feel.
Opportunity: develop more touchpoints for compassion on the platform and throughout the user experience.
Fostering trust
Patients have a perceived uniqueness of their health needs. This means they feel that AI may not adequately address their individual health requirements, leading to doubts about the technology's ability to cater to their specific needs.
As of now, the product feels very impersonal, with little confirmation and recognition of the users sentiments until the diagnosis at the end.
Opportunity: improve trust between the user and platform, through a more personalised experience.
ALIGNING ON THE PATH FORWARD
Building a healthy foundation
We started by identifying an accountability buddy at the hackathon, who took on an informal role of Product Manager. We set out a timeframe for ourselves to make sure we covered the necessary bases in the time available. We tasked our PM to keep us on track, ask us critical questions, as well as be a bouncing board for feedback.
SECONDARY RESEARCH
Assessing assumptions
I guided the group in outlining our assumptions and allocated time to explore the problem space as a team. Due to the time constraints, I decided that it would be useful to explore qualitative secondary research, as a descriptive means to understanding user frustrations and immerse ourselves into the mind of healthcare and telemedicine users.
"Patients within the healthcare space feel unseen and ‘genericised’ instead of as unique human beings with unique health needs…"
"...we all pine for a personal touch, empathy of a healthcare provider who knows our history, “gets us.”
"there is still a fundamental disconnect between my data – be it medical history or even recently received lab results – and the decision-making and context my primary care physician is using when engaging with me for any diagnoses."
"Patients within the healthcare space feel unseen and ‘genericised’ instead of as unique human beings with unique health needs…"
"...we all pine for a personal touch, empathy of a healthcare provider who knows our history, “gets us.”
"there is still a fundamental disconnect between my data – be it medical history or even recently received lab results – and the decision-making and context my primary care physician is using when engaging with me for any diagnoses."
"Patients within the healthcare space feel unseen and ‘genericised’ instead of as unique human beings with unique health needs…"
"...we all pine for a personal touch, empathy of a healthcare provider who knows our history, “gets us.”
"there is still a fundamental disconnect between my data – be it medical history or even recently received lab results – and the decision-making and context my primary care physician is using when engaging with me for any diagnoses."
"Patients within the healthcare space feel unseen and ‘genericised’ instead of as unique human beings with unique health needs…"
"...we all pine for a personal touch, empathy of a healthcare provider who knows our history, “gets us.”
"there is still a fundamental disconnect between my data – be it medical history or even recently received lab results – and the decision-making and context my primary care physician is using when engaging with me for any diagnoses."
Joanna Ruiz, Deloitte DigitalPatients don't want
To have a generalised experience
To be left guessing on what to do next when distressed
To be misdiagnosed
Patients want
To be understood and heard
Speed and convenience of service
Reassurance and guidance on next steps
PRODUCT RESEARCH
Pulse check
With an understanding of user expectations when it comes to healthcare services, and certain assumptions validated, we used this context to conduct product research to situate ourselves in their minds. This gave us perspective when evaluating the current symptom checking process that Healthily offers — currently marketed by the company as The Symptom Checker. We evaluated its features, functionalities, user interface, and user experience and identified areas of opportunity and that require further exploration.
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
SECONDARY RESEARCH
Assessing assumptions
I guided the group in outlining our assumptions and allocated time to explore the problem space as a team. Due to the time constraints, I decided that it would be useful to explore qualitative secondary research, as a descriptive means to understanding user frustrations and immerse ourselves into the mind of healthcare and telemedicine users.
"Patients within the healthcare space feel unseen and ‘genericised’ instead of as unique human beings with unique health needs…"
"...we all pine for a personal touch, empathy of a healthcare provider who knows our history, “gets us.”
"there is still a fundamental disconnect between my data – be it medical history or even recently received lab results – and the decision-making and context my primary care physician is using when engaging with me for any diagnoses."
"Patients within the healthcare space feel unseen and ‘genericised’ instead of as unique human beings with unique health needs…"
"...we all pine for a personal touch, empathy of a healthcare provider who knows our history, “gets us.”
"there is still a fundamental disconnect between my data – be it medical history or even recently received lab results – and the decision-making and context my primary care physician is using when engaging with me for any diagnoses."
"Patients within the healthcare space feel unseen and ‘genericised’ instead of as unique human beings with unique health needs…"
"...we all pine for a personal touch, empathy of a healthcare provider who knows our history, “gets us.”
"there is still a fundamental disconnect between my data – be it medical history or even recently received lab results – and the decision-making and context my primary care physician is using when engaging with me for any diagnoses."
"Patients within the healthcare space feel unseen and ‘genericised’ instead of as unique human beings with unique health needs…"
"...we all pine for a personal touch, empathy of a healthcare provider who knows our history, “gets us.”
"there is still a fundamental disconnect between my data – be it medical history or even recently received lab results – and the decision-making and context my primary care physician is using when engaging with me for any diagnoses."
Patients don't want
To have a generalised experience
To be left guessing on what to do next when distressed
To be misdiagnosed
Patients want
To be understood and heard
Speed and convenience of service
Reassurance and guidance on next steps
Designing for visual confirmation
Tasked with designing for desktop, we decided to split the screen and use the right-hand side as way to give visual confirmation to the user that the AI is processing their input.
Using visuals throughout the symptom checking process introduces a more user friendly and engaging method that not only makes them feel heard, but also helps with engagement rates.
This visual stimulation would ensure that users pay more attention throughout the process and give more accurate and "thought out" answers, rather than ones that are rushed and distracted.
Keeping users on the happy path
About halfway through the symptom checking process, questions begin getting very repetitive and similar in nature, as the algorithm begins narrowing down potential health causes. With no context for the user, this can feel unusual and give them a sense that something is "wrong," or worse — that the product and algorithm is not working.
We explored ways in which the user interface could be designed to give preemptive context to the user, so as to reduce feelings of distrust towards the platform.
Introducing more human touches
The current marketing of the product, onboarding screens as well as process on the whole is very text heavy, and is missing an opportunity to introduce reassurance through the human gaze.
Studies show that there is a higher engagement with a product when the fourth wall is broken with a user. To ensure the user does not feel alone, we searched for more positive and affirming imagery with people looking at the user directly to make them feel supported and heard.
To help them understand that the AI is not a soulless creation but one that represents the work of many people whose best interest is to make their lives better.
KEY FINDINGS AND OPPORTUNITIES
Healing the open wound
Expanding empathy
The integration of AI should not compromise essential human qualities like empathy that patients seek in their interactions with healthcare professionals.
Currently, the product lacks elements of understanding and compassion, not only throughout the symptom checking process, but also in its overall look and feel.
Opportunity: develop more touchpoints for compassion on the platform and throughout the user experience.
Fostering trust
Patients have a perceived uniqueness of their health needs. This means they feel that AI may not adequately address their individual health requirements, leading to doubts about the technology's ability to cater to their specific needs.
As of now, the product feels very impersonal, with little confirmation and recognition of the users sentiments until the diagnosis at the end.
Opportunity: improve trust between the user and platform, through a more personalised experience.
How might we support patients to access the correct healthcare channels?
PROTO-PERSONA
Angelita's journey
A Healthily user who self-identifies as a hypochondriac
Often worried and anxious about her health
Seeks reassurance from friends, family and healthcare professionals
Looking for quick answers to her health concerns
Has been misdirected by healthcare providers in the past
EXPLORATORY SKETCHES AND SOLUTIONS
Cooking up ideas
With an understanding of user needs, I led and explained ideation exercises like Crazy 8s to guide our thinking process as a team. My teammate came from a business strategy background and was unfamiliar with ideation exercises. I explained the benefits of involvements from all teammates and emphasised that ideas from both a technical and non-technical background are equally valuable in this stage of the design process.
SPOTLIGHT ONE
Making the user feel heard
What are ways in which we can make the user feel heard, when inputting data? How can machine learning validate how users feel? We explored ways in which we can design for users seeking "reassurance."
Designing for visual confirmation
Tasked with designing for desktop, we decided to split the screen and use the right-hand side as way to give visual confirmation to the user that the AI is processing their input.
Using visuals throughout the symptom checking process introduces a more user friendly and engaging method that not only makes them feel heard, but also helps with engagement rates.
This visual stimulation would ensure that users pay more attention throughout the process and give more accurate and "thought out" answers, rather than ones that are rushed and distracted.
Keeping users on the happy path
About halfway through the symptom checking process, questions begin getting very repetitive and similar in nature, as the algorithm begins narrowing down potential health causes. With no context for the user, this can feel unusual and give them a sense that something is "wrong," or worse — that the product and algorithm is not working.
We explored ways in which the user interface could be designed to give preemptive context to the user, so as to reduce feelings of distrust towards the platform.
Introducing more human touches
The current marketing of the product, onboarding screens as well as process on the whole is very text heavy, and is missing an opportunity to introduce reassurance through the human gaze.
Studies show that there is a higher engagement with a product when the fourth wall is broken with a user. To ensure the user does not feel alone, we searched for more positive and affirming imagery with people looking at the user directly to make them feel supported and heard.
To help them understand that the AI is not a soulless creation but one that represents the work of many people whose best interest is to make their lives better.
SPOTLIGHT TWO
Reimagining the product identity
We did a deep dive of the current Healthily app and noticed the notion of being your own "health detective."
Notably, current product branding of AI positions LLM's as an "assistant" on the user's journey: for example, Shopify's Sidekick (a partner in your e-commerce adventures), Microsoft's Co-Pilot (a partner that helps you navigate life and work), or Google's Gemini which is Latin for "twin."
Onboarding screens on Healthily mobile app
Defining the role of the product
While the current name of the product is Symptom Checker, we chose to repurpose and expand on the concept of "Health Detective:"
Assistant on health discoveries (not quite a "medical" expert but an expert nonetheless)
A relatable, friendly figure, who is playing on the same "team" as the user
Educated, driven and dedicated to the user's cause
Colours and tones
Studies show that green has a therapeutic and calming effect on the mind. I recognise that users like Angelita are likely to be frustrated, distressed, agitated or at least worried about their wellbeing when navigating the application.
Using the company branding documents that I sourced on Dribbble, I decided to explore different combinations of the company's pre-approved shades of green.
Shapes
When picking the Health Detective's logo, I identified secondary logos that the company has designed but seldom uses.
I decided to opt for one of these pre-approved logos as the Health Detective’s identifying shape. With semblance in look and feel to OpenAI’s ChatGPT or Perplexity's AI, it is a nod to the notion of being "generative."
The repetitive circular shapes feel both comforting and structured, as well as exponential in possibilities.
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
Feels lengthy, formal and repetitive, and lacks engaging elements
Absence of empathy and compassion in both look, feel, and tone of voice
Lack of clarity as to the digital next steps on the platform, as well as no clear call to action on the platform
DESIGN AND DEVELOPMENT
Say hello to Lily
Softens the slightly harder, more masculine and intense connotations of "detective"
Approachable and on-brand name that rolls off the tongue: Lily by Healthily
Alludes to a “lily pad,” which ties into the shapes of the logo and gives a natural, healthy and relaxing feel
DESIGN CRITIQUE
Final touches
I designed our solution in greyscale. After a design critique from the team (in the form of Popcorn discussion), it was edited a final time before being developed in high fidelity. We relied on prototyping the solution to showcase our solution to the judges.
■ Before and after
SPOTLIGHT ONE
Making the user feel heard
What are ways in which we can make the user feel heard, when inputting data? How can machine learning validate how users feel? We explored ways in which we can design for users seeking "reassurance."
EXPLORATORY SKETCHES AND SOLUTIONS
Cooking up ideas
With an understanding of user needs, I led and explained ideation exercises like Crazy 8s to guide our thinking process as a team. My teammate came from a business strategy background and was unfamiliar with ideation exercises. I explained the benefits of involvements from all teammates and emphasised that ideas from both a technical and non-technical background are equally valuable in this stage of the design process.
Defining the role of the product
While the current name of the product is Symptom Checker, we chose to repurpose and expand on the concept of "Health Detective:"
Assistant on health discoveries (not quite a "medical" expert but an expert nonetheless)
A relatable, friendly figure, who is playing on the same "team" as the user
Educated, driven and dedicated to the user's cause
Colours and tones
Studies show that green has a therapeutic and calming effect on the mind. I recognise that users like Angelita are likely to be frustrated, distressed, agitated or at least worried about their wellbeing when navigating the application.
Using the company branding documents that I sourced on Dribbble, I decided to explore different combinations of the company's pre-approved shades of green.
Shapes
When picking the Health Detective's logo, I identified secondary logos that the company has designed but seldom uses.
I decided to opt for one of these pre-approved logos as the Health Detective’s identifying shape. With semblance in look and feel to OpenAI’s ChatGPT or Perplexity's AI, it is a nod to the notion of being "generative."
The repetitive circular shapes feel both comforting and structured, as well as exponential in possibilities.
Keeping users on the happy path
About halfway through the symptom checking process, questions begin getting very repetitive and similar in nature, as the algorithm begins narrowing down potential health causes. With no context for the user, this can feel unusual and give them a sense that something is "wrong," or worse — that the product and algorithm is not working.
We explored ways in which the user interface could be designed to give preemptive context to the user, so as to reduce feelings of distrust towards the platform.
Introducing more human touches
The current marketing of the product, onboarding screens as well as process on the whole is very text heavy, and is missing an opportunity to introduce reassurance through the human gaze.
Studies show that there is a higher engagement with a product when the fourth wall is broken with a user. To ensure the user does not feel alone, we searched for more positive and affirming imagery with people looking at the user directly to make them feel supported and heard.
To help them understand that the AI is not a soulless creation but one that represents the work of many people whose best interest is to make their lives better.
Designing for visual confirmation
Tasked with designing for desktop, we decided to split the screen and use the right-hand side as way to give visual confirmation to the user that the AI is processing their input.
Using visuals throughout the symptom checking process introduces a more user friendly and engaging method that not only makes them feel heard, but also helps with engagement rates.
This visual stimulation would ensure that users pay more attention throughout the process and give more accurate and "thought out" answers, rather than ones that are rushed and distracted.
KEY FINDINGS AND OPPORTUNITIES
Healing the open wound
Expanding on empathy
The integration of AI should not compromise essential human qualities like empathy that patients seek in their interactions with healthcare professionals.
Currently, the product lacks elements of understanding and compassion, not only throughout the symptom checking process, but also in its overall look and feel.
Opportunity: develop more touchpoints for compassion on the platform and throughout the user experience.
Fostering trust
Patients have a perceived uniqueness of their health needs. This means they feel that AI may not adequately address their individual health requirements, leading to doubts about the technology's ability to cater to their specific needs.
As of now, the product feels very impersonal, with little confirmation and recognition of the users sentiments until the diagnosis at the end.
Opportunity: improve trust between the user and platform, through a more personalised experience.
How might we support patients to access the correct healthcare channels?
PROTO-PERSONA
Angelita's journey
A Healthily user who self-identifies as a hypochondriac
Often worried and anxious about her health
Seeks reassurance from friends, family and healthcare professionals
Looking for quick answers to her health concerns
Has been misdirected by healthcare providers in the past
Often worried and anxious about her health
Seeks reassurance from friends, family and healthcare professionals
Looking for quick answers to her health concerns
Has been misdirected by healthcare providers in the past
EXPLORATORY SKETCHES AND SOLUTIONS
Cooking up ideas
With an understanding of user needs, I led and explained ideation exercises like Crazy 8s to guide our thinking process as a team. My teammate came from a business strategy background and was unfamiliar with ideation exercises. I explained the benefits of involvements from all teammates and emphasised that ideas from both a technical and non-technical background are equally valuable in this stage of the design process.
USER PERSONA
Pulse check
With an understanding of user expectations when it comes to healthcare services, and certain assumptions validated, we used this context to conduct product research to situate ourselves in their minds. This gave us perspective when evaluating the current symptom checking process that Healthily offers — currently marketed by the company as The Symptom Checker. We evaluated its features, functionalities, user interface, and user experience and identified areas of opportunity and that require further exploration.
NEXT STEPS
Positive trajectory
Take time to improve the quality of our research to best align our designs with user needs, by interviewing other users: both existing Healthily users as well as others in the target market.
Consult with the data science and software engineering team at Healthily, gather input on these changes, and whether adjustments or iterations need to be made due to technical needs.
Several rounds of user testing and usability tests, prioritising feedback to iterate in the most time efficient way possible (e.g. developing back or undo buttons in the assessment flow).
Once launched, collaborate closely with the data team, using methods like A/B testing to best understand the success rate of our designs and where we can improve.
SPOTLIGHT TWO
Reimagining the product identity
We did a deep dive of the current Healthily app and noticed the notion of being your own "health detective."
Notably, current product branding of AI positions LLM's as an "assistant" on the user's journey: for example, Shopify's Sidekick (a partner in your e-commerce adventures), Microsoft's Co-Pilot (a partner that helps you navigate life and work), or Google's Gemini which is Latin for "twin."
Onboarding screens on Healthily mobile app
Defining the role of the product
While the current name of the product is Symptom Checker, we chose to repurpose and expand on the concept of "Health Detective:"
Assistant on health discoveries (not quite a "medical" expert but an expert nonetheless)
A relatable, friendly figure, who is playing on the same "team" as the user
Educated, driven and dedicated to the user's cause
Colours and tones
Studies show that green has a therapeutic and calming effect on the mind. I recognise that users like Angelita are likely to be frustrated, distressed, agitated or at least worried about their wellbeing when navigating the application.
Using the company branding documents that I sourced on Dribbble, I decided to explore different combinations of the company's pre-approved shades of green.
Shapes
When picking the Health Detective's logo, I identified secondary logos that the company has designed but seldom uses.
I decided to opt for one of these pre-approved logos as the Health Detective’s identifying shape. With semblance in look and feel to OpenAI’s ChatGPT or Perplexity's AI, it is a nod to the notion of being "generative."
The repetitive circular shapes feel both comforting and structured, as well as exponential in possibilities.
SPOTLIGHT TWO
Reimagining the product identity
We did a deep dive of the current Healthily app and noticed the notion of being your own "health detective."
Notably, current product branding of AI positions LLM's as an "assistant" on the user's journey: for example, Shopify's Sidekick (a partner in your e-commerce adventures), Microsoft's Co-Pilot (a partner that helps you navigate life and work), or Google's Gemini which is Latin for "twin."
Onboarding screens on
Healthily mobile app
DESIGN AND DEVELOPMENT
Say hello to Lily
Softens the slightly harder, more masculine and intense connotations of "detective"
Approachable and on-brand name that rolls off the tongue: Lily by Healthily
Alludes to a “lily pad,” which ties into the shapes of the logo and gives a natural, healthy and relaxing feel
DESIGN CRITIQUE
Final touches
I designed our solution in greyscale. After a design critique from the team (in the form of Popcorn discussion), it was edited a final time before being developed in high fidelity. We relied on prototyping the solution to showcase our solution to the judges.
DESIGN AND DEVELOPMENT
Say hello to Lily
Softens the slightly harder, more masculine and intense connotations of "detective"
Approachable and on-brand name that rolls off the tongue: Lily by Healthily
Alludes to a “lily pad,” which ties into the shapes of the logo and gives a natural, healthy and relaxing feel
DESIGN CRITIQUE
Final touches
I designed our solution in greyscale. After a design critique from the team (in the form of Popcorn discussion), it was edited a final time before being developed in high fidelity. We relied on prototyping the solution to showcase our solution to the judges.
■ Before and after
Desktop view
Website
ORIGINAL
Worrying imagery
Cold colours
Plain language
NEW VERSION
Soothing imagery
Personalised feel
Warm and nurturing tone
Onboarding
ORIGINAL
No onboarding screens
No guidance on what to expect
Feel the lack of support
NEW VERSION
Comforting, supportive, human-focused experience
Recovery-focused imagery and content
Clear expectations set
Assessment
ORIGINAL
Repetitive content and layout
Lack of engaging elements
Plain colours and shapes
NEW VERSION
Introducing more interactive and visual elements related to progress and completion (e.g. iconography, clearer labelling, interactive assets)
Supportive, relaxed conversational look and feel
Final report
ORIGINAL
Unclear platform copy in relation to the final call to action
Lack of guidance on digital next steps and benefits of Lily
NEW VERSION
Clear copy for the final call to action
Guidance on digital next steps and benefits of Lily
Tone
ORIGINAL
Lack of empathy and compassion in language, tone and assets
NEW VERSION
Humanised understanding and supportive language around expectations
Desktop view
Website
ORIGINAL
Worrying imagery
Cold colours
Plain language
NEW VERSION
Soothing imagery
Personalised feel
Warm and nurturing tone
Onboarding
ORIGINAL
No onboarding screens
No guidance on what to expect
Feel the lack of support
NEW VERSION
Comforting, supportive, human-focused experience
Recovery-focused imagery and content
Clear expectations set
Assessment
ORIGINAL
Repetitive content and layout
Lack of engaging elements
Plain colours and shapes
NEW VERSION
Introducing more interactive and visual elements related to progress and completion (e.g. iconography, clearer labelling, interactive assets)
Supportive, relaxed conversational look and feel
Tone
ORIGINAL
Lack of empathy and compassion in language, tone, colours, assets, etc.
NEW VERSION
Humanised understanding and supportive language around expectations
Final report
ORIGINAL
Unclear platform copy in relation to the final call to action
Lack of guidance on digital next steps and benefits of Lily
NEW VERSION
Clear copy for the final call to action
Guidance on digital next steps and benefits of Lily
Desktop view
Website
ORIGINAL
Worrying imagery
Cold colours
Plain language
NEW VERSION
Soothing imagery
Personalised feel
Warm and nurturing tone
Onboarding
ORIGINAL
No onboarding screens
No guidance on what to expect
Feel the lack of support
NEW VERSION
Comforting, supportive, human-focused experience
Recovery-focused imagery and content
Clear expectations set
Assessment
ORIGINAL
Repetitive content and layout
Lack of engaging elements
Plain colours and shapes
NEW VERSION
Introducing more interactive and visual elements related to progress and completion (e.g. iconography, clearer labelling, interactive assets)
Supportive, relaxed conversational look and feel
Tone
ORIGINAL
Lack of empathy and compassion in language, tone, colours, assets, etc.
NEW VERSION
Humanised understanding and supportive language around expectations
Final report
ORIGINAL
Unclear platform copy in relation to the final call to action
Lack of guidance on digital next steps and benefits of Lily
NEW VERSION
Clear copy for the final call to action
Guidance on digital next steps and benefits of Lily
NEXT STEPS
Positive trajectory
Take time to improve the quality of our research to best align our designs with user needs, by interviewing other users: both existing Healthily users as well as others in the target market.
Consult with the data science and software engineering team at Healthily, gather input on these changes, and whether adjustments or iterations need to be made due to technical needs.
Several rounds of user testing and usability tests, prioritising feedback to iterate in the most time efficient way possible (e.g. developing back or undo buttons in the assessment flow).
Once launched, collaborate closely with the data team, using methods like A/B testing to best understand the success rate of our designs and where we can improve.
■ Conclusion
■ Conclusion
■ Before and after
Representing the team and explaining our design thinking
PRESENTING THE PROJECT
Clean bill of health
I developed a live demo of the Figma prototype and recorded the presentation with Loom for the judges at Encode and Healthily.
We were awarded winners of the Encode hackathon for the Healthily problem space. Our team of two, with a no-code, design-led solution secured the victory.
Representing the team and explaining our design thinking
LESSONS FOR MY NEXT HACKATHON
Future prescriptions
NEXT STEPS
Positive trajectory
LESSONS FOR MY NEXT HACKATHON
Future prescriptions
Take time to improve the quality of our research to best align our designs with user needs, by interviewing other users: both existing Healthily users as well as others in the target market.
Consult with the data science and software engineering team at Healthily, gather input on these changes, and whether adjustments or iterations need to be made due to technical needs.
Several rounds of user testing and usability tests, prioritising feedback to iterate in the most time efficient way possible (e.g. developing back or undo buttons in the assessment flow).
Once launched, collaborate closely with the data team, using methods like A/B testing to best understand the success rate of our designs and where we can improve.
LESSONS FOR MY NEXT HACKATHON
Future prescriptions
Open communication
My teammate and I were complete strangers at the beginning of the event and became good friends by the end of it. Before starting, it was helpful to outline what our communication styles, as well as our strengths and weaknesses.
About 15 minutes in, he said, "I live by the rule of: do what you say you are going to do." This open exchange at the start allowed us to not only understand each other very early on, but also trust in each other deeply. A 72 hour hackathon can feel very much like doing a trust fall with a stranger. The presence or absence of that chat at the start can make or break a team.
The power of pacing yourself
As other teams rushed into their design solutions, we realised it was a marathon not a sprint very early on. Although it was tempting to fall into the pattern of other teams, we made sure to get a good night's sleep on the first night. When designing without a solution in mind (in the exploratory phase) in such a short time span, it's easy to feel panicked, stay up late and rush into a solution. Avoid burnout by trusting the UX process and spread your energies across the entire 72 hours as evenly as possible. Don't worry, the adrenaline will definitely be enough to keep you awake on the last night!
I did this well!
Make room for error
We left only a few minutes for the final hand-in, and this sleep-deprived rush meant that we didn't hand over the documents with the correct viewing permissions. Luckily, I noticed 30 minutes after the deadline, and we were given permission to resubmit the files. However, other competitions might not be as lenient — leaving room for these final checks is crucial.
Let's reconvene in an hour
The team was very small, and the time was very short. We navigated this by leaving room for quick and impromptu feedback as well as having scheduled time for longer check-ins with each other every 2-3 hours.
This scheduled time was helpful as it allowed us to ask each other questions in bulk rather than ask each other continuously for opinions.
I did this well!
