Flick AI
AI-powered chat assistant that creates personalized, task-focused interfaces for older adults.
trophy
Top 5 Finalist
2024 Cheng-Wu Innovation Challenge
Indiana University Shoemaker Lab
Project overview
Flick is a conversational assistant that creates personalized, task-focused interfaces, enabling older adults to complete online tasks independently without learning multiple applications and websites.
Our team was formed on the vision of democratizing technology to make digital ecosystems inclusive for all. We saw our friends and family struggle to use technology for doing elementary tasks, showing that many people find it hard to utilize the digital tools they own for their everyday purposes.
Timeline
3 months (Mar-May 2023)
Role
Product Designer
Team
2 Product Designers
1 AI Engineer
Tools
Figma
Platform
iOS app
Status
Shipped
Contribution
UI design
Prototypes
My contribution
My role involved conceptualizing and designing the app's interface. I created all the high-fidelity designs and prototypes and contributed to research by conducting interviews and user testing sessions.
Problem context
Even though older adults have access to the latest technology, many find it challenging to use effectively. Everyday tasks like ordering groceries, booking a cab, or paying bills online can be difficult for them.
Meet Susan
Susan is a single mother in her mid-50s. She values her independence but isn't very comfortable with technology. Recently, she moved into a new apartment and decides to change her address with Chase.
Here’s how many links and pages Susan would have to navigate through in the Chase app just to change her address:
Above: Existing Chase app flow to change primary address.
Solution highlights
Here's how Susan changed her address with Chase using Flick.
She types the task in her own words, and Flick recognizes the actionable parameters.
Flick asks a follow-up question to refine the task, ensuring her intent is clearly understood.
She chooses how to authorize Flick to access her account.
Once logged in, Flick bypasses navigation hurdles and presents Susan with her addresses.
Susan makes the necessary updates, and she's all set!
Research insights
Speaking with older adults in our friends and family, we found that most current apps, despite being designed for everyone, fail to cater to the specific needs of Gen X (45-60 years old).
They impose a high cognitive load.
They're hard to navigate.
And they usually have steep learning curves.
Our research showed that 67% of Gen X relies on their children or trusted individuals for tech assistance.
I'm sure we've all tried to teach our parents how to change their Gmail password or order something online. It's not always helpful. These is mainly because of 3 main reasons:
support
It increases dependency and reduces self-confidence.
psychology_alt
There's a generational gap in teaching and learning.
phone_disabled
It's harder to help when people aren't physically close.
Although platforms like Senior Planet and Candoo provide tech support for older adults, they often fall short.
stack_hexagon
These platforms don't accommodate different learning styles.
task
Not everyone wants to learn. Some just want to get their tasks done.
indeterminate_question_box
Ironically, these platforms are complex to use in and of themselves.
Core challenge
How might we empower older adults to use technology effectively and independently without having to learn it?
After all, technology should adapt to the users instead of expecting them to adapt to it.
User flows
Since this was a novel concept at the time, we spent considerable time brainstorming the app's functionality, aiming to balance technical feasibility, simplicity, and user intuitiveness.
Creating user flows was immensely helpful, providing a big-picture understanding of how things would work and allowing us to focus on different aspects of the flow in detail.
After multiple iterations, we finalized the following high-level user flow for the app's functionality. Through cross-collaboration between designers and the AI engineer, we ensured each step was backed by strong technical and logical reasoning.
Once we had a solid foundation, we decided to integrate a real-life use case to demonstrate how a user would interact with Flick. We chose the relatively complex scenario of downloading a tax form to address as many potential loopholes as possible.
Visual exploration & iterations
The next phase, one of the most critical stages of the project, involved designing a visual interface that was focused, minimal, effective, and easy to understand and interact with.
I began with the most basic sketches on a blank canvas and iterated continuously, gradually adding visual complexity and polish to the designs. Here’s a sequence of iterations showcasing the visualization process.
Basic pen-and-paper sketches to capture my initial ideas on canvas.
Initial wireframes with some real data.
Added details to the response cards, one of the key components of the interface.
Continued refining the design for added polish and began experimenting with animations.
I also explored a high-contrast brutalist design approach to enhance accessibility and align with current trends.
Final designs & key parts of the solution
Introducing Flick - an application designed to dynamically generate interfaces that align with user intent.
Understanding user intent and identifying actionable parameters
Flick can adapt to the user's unique way of speaking and understanding. This helps reduce cognitive load by not having to navigate complex interfaces or remember specific terms.
question_mark
Why type when you can speak? Wouldn't that be easier?
Research revealed that most people are uncomfortable speaking aloud about their actions when others are nearby, which is often the case. It's a major invasion of their privacy.
Ensuring clarity and relevance
Follow-up questions clarify vague prompts to create relevant interfaces. This simplifies the experience by avoiding irrelevant details and reduces the chance of mistakes.
Enhancing usability through reachability
Placing primary actions in the 'green' zone keeps them within the user's most comfortable reach, especially for those with limited dexterity.
Letting users control important privacy choices that can't be reversed
This control reassures older adults that their privacy is protected, building trust and confidence in their interactions.
Bypassing navigation hurdles and eliminating redundancies
This saves time and reduces the cognitive load for older adults.
Improving focus and reducing overwhelm
I broke long scrolls into smaller, focused pages helps users concentrate on one task at a time. This prevents them from feeling overwhelmed by too much information at once.
Let's connect!
Testing & impact
We performed task-based testing with GenX individuals and gathered several actionable insights.
login
Onboarding older adults on the app is an unaddressed issue.
step_into
The application can be scoped down to included specific task types and services.
gpp_maybe
Privacy concerns with AI still prevail.
settings_backup_restore
Users need options to correct mistakes or change course mid-task.
Current stage