ThredA shared travel-memory app for couples.
Couples travel together and still experience the day apart. Thred records each partner's path and photos and weaves them into one shared story.
With two teammates I researched tourists in Heraklion, guerrilla-tested an AR concept with couples, and pivoted when people liked the idea more than the product.
I led field research and interviews, built the persona and journey map, designed the interaction model in Figma, built a working live map with Claude Code, and ran usability testing.
Full case study (PDF) ↗How it unfolded
Final iteration
DayoutA mobile app for planning your day from community-made activities.
Dayout lets people build a day from community-made activities and 24-hour plans. I joined as a UX research and design intern.
I interview experienced users and define the events behind the app's Amplitude funnels, which showed drop-off in onboarding and sign-up.
I write happy paths, run usability tests, and design new screens: onboarding, the home screen, and an AI create mode, now in development.
Usability testing
Three scenario tasks, each with an ideal path, then watching where real behavior diverged.
Screens
The Elite ImpactHow private aviation contributes to climate change, and what regulation could change.
The 2019 UN Climate Action Summit made climate change feel urgent. A study on celebrity private-jet emissions made it specific: the wealthiest emit far beyond their share, and private aviation is the clearest example. I made it my year-long capstone.
I wrote the research paper, scoped it with my mentor, collected a year of public flight-tracking data, built forecasting models, and shipped an interactive website, presented to a panel of advisors and industry experts.
Each targets private aviation directly, with a real-world precedent to model against.
Restrict the shortest flights
Flights under an hour almost always have a train or car alternative. France's short-haul ban was the precedent.
Cap annual emissions per person
A yearly CO₂ budget for the heaviest individual emitters, so private travel counts against a limit like everything else.
Limit carbon-credit sales
Offsets let a flight read as neutral on paper. Limiting them keeps the real emissions in view.
How I ran it
Scoped with a mentor
The paper covered four drivers of emissions. By Sprint 5 we cut to private aviation only, with three regulations to model: carbon-credit sales, flight length, and annual emissions per person.
Run with Scrum
Seven sprints, each with a backlog, a retrospective, and a mentor review. The retrospectives changed direction: less website, more data.
Public data, compiled by hand
Flight records from ADS-B Exchange and public jet-tracker accounts, pulled as JSON and compiled into per-aircraft datasets with routes, durations, fuel, and emissions.
Pacing the scope
I wanted water and electricity overconsumption too, then found that data isn't public. Narrowing to aviation is what let the project finish.
Timeboxing the work
I lost time polishing the website while the models waited. Blocking capstone hours on a schedule kept me on what mattered.
Tools chosen by outcome
Planned on Python, moved forecasting to Google Vertex AI, fell back to Excel when training took hours per model. Charts and maps in Looker Studio, embedded in the site.