Introduction
Radical AI is a collaborative lab with a global mission to advance the field of artificial intelligence for social good, particularly in creating Virtual Agents. Our team worked on developing a feature for agent Kai, an open-source AI Teaching Assistant who helps with educational tasks, making them more efficient and accessible. Our task was to create a reminder feature for Kai, which helps educators better organize their tasks, avoid missing important deadlines, customize and prioritize their responsibilities.
Stakeholders
Savitha Hayavadana(project manager)
Rosemary Paredes(design lab manager)
The team
Ahmad Assabbagh
Advait G
Casey Anno
Eric Khuu
Dylan Dinh
Sindhu Jois
My role
Competitive analysis
Wireframes
visual design
Prototyping
Design solution overview
Our primary focus for this feature was to create an intuitive and nonintrusive notification system that allows users to customize notification preferences, Design contextually relevant and personalized notifications, Enhance user retention by providing timely updates, and manage their tasks to reduce information overload.
Problem overview
For the product discovery stages, the stakeholders asked us to design a reminder feature for Kai, the AI assistant, so educators can be notified about their tasks. Based on what they told us, we formulated our problem statement.
Deep dive into the process
Paper research
From our understanding of Kai objectives and goals from the stakeholders and what they want for a reminder feature for an AI product ,We looked at peer reviewed papers by researchers from UC Irine ,Microsoft, and Carnegie Melon that included ai interfaces and intelligent agent in relation to productivity tools .
Screenshot of research papers from UC Irine ,Microsoft, and Carnegie Melon
From those research papers we concluded the major findings :
Finding1
Users held importance in having intelligent agent like Kai having human characteristics or anthropomorphism.
Finding 2
They value the ability to personalize reminder and they needed time based reminder to recall future tasks.
Finding 3
Users prefer an initial system-determined reminder setup and having flexibility to adjust reminders later is important.
Finding 4
Users prefer task assistants with a human touch and emotional intelligence.
Interview
In order to fill any gaps in the existing research , we created an interview script that can help us with lived experiences with reminder systems and any frustration or positive experience they have had.
key insight 1
Users require the ability to control and customize their reminders to fit their preferences.
key insight 2
Users need urgent notifications that effectively capture their attention.
key insight 3
Users expect a consistent user interface that aligns with the design patterns of other tools they are familiar with.
A sample of the interview script
We aimed to understand how educators manage their tasks, what reminders they use, and how they organize them. The interviews revealed what they seek in a reminder system and we were able to extract the main key insights.
Prototypes
After choosing the best version for the wireframes based on AB testing, we prototyped the wireframes to present them to stakeholders and show how the educators would set up reminders and manage their tasks to ensure that the design aligns with business goals, user needs and their expectations in the following :
Task notification reminder
Users will receive a reminder notification in the app, allowing them to snooze, close, or check off a task once it's completed.
Customization for reminders
Users can create reminders, set the date and time, and customize them based on priority.
Rewrite or Suggest with Kai
The users are able to ask Kai if he can rewrite the description for him or he can suggest a solution based on the title reminder to ease information process.
Setting up notification
Users can choose to receive notifications through the desktop version or via email.
After confirming the suitable version of the reminder feature prototype based on approval of the stakeholders we did a meeting with the UI checker and the design system designer to ensure that UI elements and the components of our design are used consistently throughout the product, and to ensures that the design adheres to accessibility standards, brand guidelines, and other design principles and what sort of components , elements and icons should be added to the design system
Design system & UI checker
Reflection & Nextsteps
always think about how to present my designs to stakeholders, rather than striving for perfection, especially in the discovery stages. My focus is on execution and clearly explaining my design decisions. Have conversations with other design teams to understand how our features will interact with one another.
Next steps for the project?
Ensuring the effectiveness of changes through further testing is key to meeting user needs on YouTube. This includes integrating redesigns with existing tools and exploring new features desired by users. Additionally, enhancing user experience across various YouTube sections remains a priority.