TripPilot: AI-Powered Travel Planning
TripPilot is an AI-powered travel planning platform designed to eliminate decision fatigue. By replacing fragmented research with an intelligent, adaptive itinerary and real-time conversational refinement, TripPilot turns 20 hours of trip planning into a 15-minute collaborative experience.
My role
Builder
Timeline
January to March 2026
My team
Just me!
Problem Space & Opportunity
Americans spend billions of hours planning trips every year.
Planning a trip can mean hours of searching across flights, hotels, maps, reviews, social media, and travel guides. I saw an opportunity to bring that fragmented process together and reduce the amount of work travelers had to do themselves.
2.4 billion
hours of excess trip-planning time each year*
Estimated annual U.S. impact based on my problem-sizing model.
Market Focus & Strategic Positioning
But not every traveler needed the same kind of help.
Rather than designing for everyone who takes a trip, I focused on who spends the most time planning, where existing tools fall short, and which travelers would get the most value from an AI planning assistant.
Competitive Landscape
Existing tools helped with pieces of the trip, but travelers still had to put everything together.
I compared the major ways people currently plan travel and found that most products focused on one part of the process: searching, booking, organizing, or generating recommendations.
Product Direction
That led me to focus TripPilot on the planning itself.
Instead of simply generating an itinerary, TripPilot would learn what someone wanted from their trip, create a plan around those preferences, and let them keep changing it through conversation.
Product Direction
I translated that strategy into 11 requirements for what TripPilot needed to do.
Understand the traveler
Preferences, budget, dates, constraints
Build the trip
Activities, timing, geography, reservations
Adapt the plan
Conversational changes and real-time refinement
The Solution
TripPilot turns hours of fragmented research into one collaborative planning experience.
Instead of treating inventory management as one universal workflow, we designed tools around the two environments where staff use them: Warehouse Intake and the Retail Floor.
How it works
Travelers start with their preferences, then build and refine the trip with AI.
01
Tell TripPilot what you're looking for.
X
02
Get a complete itinerary built around those preferences.
X
03
Ask for changes without starting over.
X
The Strategy
I built the concept around a ~$4.5 billion U.S. market opportunity.
X
Constant interruptions
Staff regularly switched between customers, donations, pricing, and inventory.
Staff workarounds
Employees had developed workarounds to compensate for gaps in the system.
Different needs
Warehouse intake and retail inventory management required different workflows.
The workflow itself needed attention.
Staff regularly switched between customers, donations, pricing, and inventory.
Constraints
From the beginning, Construction Junction asked us to create something realistic.
We kept coming back to that word as we explored the solution space. Any solution had to work within their budget, existing technology, and the day-to-day habits staff had built over years of using Reuse Retail.
Our research also challenged the assumption that every inefficient workflow needed to be replaced. Staff had developed workarounds they knew well, and changing too much at once would require them to relearn processes they could already navigate. We focused the redesign on the parts of the system where change would have the greatest value: improving accessibility, reducing unnecessary cognitive load, and making inventory data more consistent.
That last piece mattered beyond the interface. Inconsistent item names and pricing practices made the data flowing into Salesforce difficult to analyze, limiting CJ's ability to understand its inventory and make informed business decisions. The redesign gave us an opportunity to introduce more structure without disrupting the workflows staff depended on.
Familiar workflows were hard to change
Staff relied on years of workarounds, and both management and staff were hesitant to adopt major changes or AI.
The solution had to be practical
Limited funding, staffing, and technical capacity meant anything we designed needed to be affordable, maintainable, and quick to put into practice.
CJ couldn't trust its inventory data
Inconsistent records and decade-old “ghost inventory” made Salesforce unreliable.
The Outcome
Our work helped secure $30,000 in funding to move the solutions toward implementation, giving Construction Junction a path to continue the work after our team stepped away.
Takeaways
Learning when to keep pushing and when to move forward.
Capstone taught me that doing good work on a team requires more than having high standards. I care deeply about the quality of what I produce, and my instinct is often to keep refining until I feel confident in it. Over seven months, I learned that sometimes the better decision is to align, make the decision, and keep the team moving.
My role also changed throughout the project. I started as Product Manager and later moved into a more hands-on design and prototyping role, but I found myself moving between the two whenever the team needed it. That experience taught me that leadership doesn't always come from a title.