Back

SkyAssistAI

An agentic AI concept that helps airline customer support teams automate routine requests while escalating complex decisions to human agents.

My role

Product Strategy & Design

Timeline

August to December 2024

My team

Caroline Shi
Glenda Tan
Janice Lu
Peace Im
Priscilla Park

🏆 Most Likely to Invest

Voted most investable concept in CMU's Design of AI Products course.

The Opportunity

Airline support is full of repetitive work, but not every decision should be automated.

Airline support agents spend time navigating routine requests, fragmented information, and complicated passenger needs. We saw an opportunity to use agentic AI to handle repetitive tasks while keeping human judgment at the center of consequential decisions.

The Solution

An AI copilot that knows when to ask for help.

The system is designed around different levels of autonomy. Straightforward requests can be handled by the AI, while complex or high-risk situations are escalated with the relevant context and possible resolution paths already prepared.

Product Thinking

We chose human oversight over full automation.

We started with 12 concepts and evaluated them against feasibility, user value, and risk. Rather than building a fully autonomous support agent, we focused on an AI-assisted workflow that could reduce repetitive work without removing human accountability.

Three principles guided us…

Customer Control

Passengers can access human support when needed.

Human Oversight

Agents review consequential decisions before executing.

System Visibility

Agents can see AI recommendations and supporting context.

The Outcome

🏆 Most Likely to Invest

SkyAssistAI was voted “Most Likely to Invest” in CMU's Design of AI Products course, with a projected 154,000 support hours saved annually.

Takeaways

Agentic AI requires designing more boundaries than I expected.

This project pushed me to think beyond what an AI system could automate and consider where autonomy was appropriate. Giving the system more agency also meant thinking more carefully about escalation, oversight, and failure. We also discussed what this meant for human workers who might lose jobs because of this technology. While this project was speculative, it really opened my eyes to how carefully organizations need to think about bringing AI into their spaces.

The Solution

An AI copilot that knows when to ask for help.

The system is designed around different levels of autonomy. Straightforward requests can be handled by the AI, while complex or high-risk situations are escalated with the relevant context and possible resolution paths already prepared.

Takeaways

Agentic AI requires designing more boundaries than I expected.

This project pushed me to think beyond what an AI system could automate and consider where autonomy was appropriate. Giving the system more agency also meant thinking more carefully about escalation, oversight, and failure. We also discussed what this meant for human workers who might lose jobs because of this technology. While this project was speculative, it really opened my eyes to how carefully organizations need to think about bringing AI into their spaces.

Takeaways

Agentic AI requires designing more boundaries than I expected.

This project pushed me to think beyond what an AI system could automate and consider where autonomy was appropriate. Giving the system more agency also meant thinking more carefully about escalation, oversight, and failure. We also discussed what this meant for human workers who might lose jobs because of this technology. While this project was speculative, it really opened my eyes to how carefully organizations need to think about bringing AI into their spaces.

Close-up portrait of a person

There's still so much I'm excited to learn.

I'm always looking for new problems to solve and great people to learn from. Let's talk!

Let's connect!

Email me

or get in touch on

©2026 Elise Zur

Thursday, 10/29/2026