Client: Delta
Team: Pratt Institute Conversational Design Students (Sarah Geiz and Conner Meek)
Role: Product Ideation, Flow diagram, Sample scripts, Voice demos, Presentation
Timeline: 3 months (March - May 2025)
The Challenge
Delta Airlines is the world's largest airline by revenue ($61.6B in 2024) and serves over 200 million passengers annually.
Many travelers must make connections between flights in large, unfamiliar airports, often with limited time or support.
Yet, Delta, like other major US airlines, lacks convenient support for passengers managing their layover experiences, including:
- Finding unfamiliar gates quickly
- Handling last-minute changes
- Balancing time-sensitive choices like food, restrooms, or lounges
The Solution
The Delta Layover Assistant is a context-aware, voice-activated support tool within the Fly Delta app.
Travelers can initiate it during a layover to:
- Navigate to gates with real-time directions
- Get alerts for gate changes, delays, and timing
- Explore nearby amenities during longer layovers
- Escalate to live agents when needed
Designed for hands-free use with headphones or voice prompts, the assistant balances warmth and professionalism to deliver value during time-sensitive moments.
App screen (left) and notification (right) upon arrival to layover airport
The Process
The Current Landscape
The FlyDelta app already offers a chatbot designed to help users manage their travel, but it focuses on pre-flight changes like cancellations and baggage. This chatbot is tucked away in the menu under the contact option.
Our feature builds directly into the existing Fly Delta app and will be available as soon as the app is opened. It adds a voice-guided layer to the chatbot.
Current FlyDelta Messaging
Our Product: The Chatbot Input and Output
Conversational Fit
We took Google's Conversational Fit Quiz to assess how well our product will work as a conversational interface. Our layover assistant is:
- Intuitive: Mimics familiar interactions with support teams and GPS services.
- Time & Effort Saving: Helps users efficiently navigate airports, especially during tight connections.
- Multitasking Support: Offers both text and voice options for hands-free or noisy environments.
- Comfortable & Private: Addresses potential user shyness about sharing information aloud with the text-based chat feature option
- Time & Effort Saving: Helps users efficiently navigate airports, especially during tight connections.
- Multitasking Support: Offers both text and voice options for hands-free or noisy environments.
- Comfortable & Private: Addresses potential user shyness about sharing information aloud with the text-based chat feature option
Conversational Fit Quiz
Personality
The goal was to create trust in a stressful situation while allowing the traveler to stay in control.
The Core Flow: Get Me to My Gate
The flow diagram on the right shows how the assistant guides a user who arrived in their connecting airport to their next gate. The assistant:
- Is multimodal based on user preferences - on screen interactions are red, voice interactions are blue
- Uses location services for real-time navigation (can also navigate with confirmed flight info)
- Detects gate change and reroutes
- Gives options to navigate to amenities in the airport
Flow Diagram
Sample Dialogues and Demos
Sample 1: Gate Navigation
Sample 2: Long Layover Recommendations
Want to see more? View additional scripts here →
Delivering Insights
Delivery
Client Feedback
1) Positive feedback surrounding bot personality and the focus on travelers, along with the key behaviors.
2) Reassurance, discourse markers, and guidance were noted as strong features of the bot.
Future Considerations
1) Develop the multimodal aspects of the bot, including the maps on the screen.
2) Build out the flow diagram to include the variety of factors that can affect the conversation.
3) Map out a user journey map bringing this feature into the existing FlyDelta app.
Thanks for going through my case study!
