KAI.AI Calendar

Simplifying complex scheduling through conversational AI and human-centered design

MY ROLE

UX Design, User research, Design Synthesis, Prototyping

TEAM

5 Designers

5 Designers

TIMELINE

5 Weeks

TOOLS

Figma

PROJECT OVERVIEW

KAI.AI is an AI-assistive tool that helps educators generate lesson plans and classroom materials via a chatbot. I led the redesign of the Calendar & Scheduling Ecosystem, moving the product from an initial concept to a refined, conversational planning interface.

How can we improve this page?

usability testing

Method: Moderated usability study with 4 K-12 teachers. Conducted a "stress test" where the teachers were asked to complete and manage a filled out reminders page. The goal was to discover friction in V1.

Context switching caused frustration

Context switching caused frustration

Forcing teachers to constantly switch contexts between generating materials in the chat and manually mapping them onto a separate calendar created a heavy cognitive tax.

Forcing teachers to constantly switch contexts between generating materials in the chat and manually mapping them onto a separate calendar created a heavy cognitive tax.

The Burden of Manual Verification

The Burden of Manual Verification

Teachers having to manually check off a calendar event felt like an extra administrative chore, negating the promised benefits of an automated assistant.

Teachers having to manually check off a calendar event felt like an extra administrative chore, negating the promised benefits of an automated assistant.

Screen was too busy

Screen was too busy

We noticed the teachers spent a significant amount of time looking for reminders on this main page. There were too many panels with different things going on.

We noticed the teachers spent a significant amount of time looking for reminders on this main page. There were too many panels with different things going on.

What did we learn?

  • Educators faced a high interaction cost.

  • Fragmented information architecture increased cognitive load.

User Journey

Redesign Approach

Transform a reminders-centric system into a calendar-first, AI-supported planning experience

I led the initiative to shift the product from a reminders-centric system to a calendar-first planning experience by removing reminders entirely and establishing the calendar as the single source of truth. We then integrated AI chat directly into the calendar, allowing educators to create and edit events conversationally while retaining full manual control. This approach reduced complexity and better aligned the experience with how educators naturally plan their schedules.

Design Solutions

Reminders → Calendar

I removed the redundant reminders system to resolve visual competition between tasks and scheduled events. This streamlined architecture improves information scannability and eliminates the busywork of manually marking repetitive tasks as complete.

Reducing Context Switching

By placing the AI alongside the schedule, educators can now prompt, verify, and edit their day in a single, unified workspace. This integration allows for a two-way sync where conversational prompts immediately reflect on the calendar grid. On the main chat page, CTA buttons lead users to the calendar.

Adding ai panel into the calendar

Main chat with CTA to calendar

KAI.AI Calendar Showcase

Reflection

Empowering Users by Removing Friction

This project reinforced that strong AI design isn't about maximizing automation, but about strategically preserving user agency. By stripping away the automated reminders that created digital busywork, I learned that subtracting features can sometimes yield a far more empowering and frictionless user experience than adding them.

What's Next?

Although KAI.AI has rebranded into Reality.AI and the company has moved on to different projects, the future of this project could be explored more through expanding AI actions, and additional usability testing should be done to validate our design decisions. Actions such as AI calendar suggestions and generating weekly summaries would deepen the value of the AI beyond simple commands.

Thank you for reading!

Explore more work