I bring spatial reasoning to UX design, bridging the physical spaces people move through and the digital ones they use every day.
Class projects, studio work, and the art I make outside of design.
The home page of a recipe app for easy, healthy recipes, with a Healthy mode toggle and a groceries tool that suggests meals from what you already have. This was mostly a UI design project, built on Gestalt principles and a 4-column grid.
A website that helps college students find and join study groups. I took it from content maps and lo-fi wireframes to a high-fidelity design.
An environmental middle school in Lawrenceville, Pittsburgh, that uses rewilding to help children explore ecology while opening green space to nearby families. Designed with my partner Marra Storey in Poiesis III, Fall 2025.
First floor plan and Section A
A door painted with my loved ones that opens onto a black hole, about the things we all keep locked away. At my school, people anonymously wrote down what they hide so the black hole could carry it away.
A self-portrait where tearing away the current me reveals the free, younger me underneath, grinning even dressed as a clown.
Acrylic paint and lino print showing a keyhole view into ourselves, with all our emotions mixed together.
A pencil sketch of everyday objects. Not perfect, but just right.
An e-book aggregator concept that helps Indian digital readers find any book, in the format and language they need, across a fragmented set of platforms.
Readers know which book they want but can't tell which platform carries it, so they give up or turn to piracy.
Secondary research, a 50-person survey, 4 stakeholder interviews, empathy maps, and 2 personas.
A unified search concept with price, format, and language comparison, backed by an IA, task flows, wireframes, and a high-fidelity discovery page.
Indian digital readers struggle to discover and access e-books across India’s fragmented platform ecosystem. They spend excessive time searching multiple platforms, often resorting to piracy because they do not know which platform carries the book they want or in the format they need.
Lack of discoverability within India’s fragmented e-book ecosystem.
Indian readers with access to the internet and devices.
I started with secondary research to understand the reality of the market in India and worldwide, then moved to surveys and interviews for first-person experiences. Combining the two led to empathy maps and personas.
The Indian e-book market is growing fast. E-books made up 15% of total book sales in India in 2022, up from 5% in 2018, and the market is expected to grow at more than 6% CAGR from 2024 to 2029. Behavior matters more than demographics: purchase intent and brand engagement predict e-book buying better than income or age.
Source: Debnath (2025), from ISBN metadata. Publishing jumped in 2020 and kept growing after the pandemic instead of levelling off.
A survey of 50 people (68% aged 18 to 24, 88% from North India) showed that the books people want usually exist digitally. Finding them is the hard part.
I then ran four interviews, each with a different stakeholder: a reader who prefers e-books, a reader who prefers paperbacks, a physical bookstore (Bahrisons), and a reader who uses piracy. These narrowed the audience to two groups: the Digital Reader and the Piracy Digital Reader.
Book Stop is an e-book aggregator that brings together the fragmented Indian ecosystem (Kindle, Google Play, Apple Books, and regional sources) and offers:
Main navigation covers search, home, library, friends, account, and a help center. Search includes a chatbot, each book page compares prices, formats, and languages, and the library holds saved books, a wishlist, and reading history.
Three to-be flows: discovering a book, first-time onboarding, and finding friends.
I sketched the search, discovery, and book pages on paper first, exploring a carousel for recommendations and rows for trending books and friends' picks. I then turned the discovery page into a first low-fidelity wireframe.
This is a high-fidelity design of the Book Stop discovery page, where readers start their search. It brings the research into one screen: a single search across every platform with format filters, picks matched to your reading with the platform and price shown upfront, top 10 lists by category, and books your friends are saving.
The research showed that readers don't need more books. They need a clearer path to the ones that already exist. The low and high fidelity wireframes are a first attempt and require more iteration on my part, and I'll keep developing Book Stop as I learn more about UX design.
An adaptive interface for the Transit app that learns your daily routine and reduces friction points in your everyday commute by reducing the number of screens and predicting possible interferences.
Routine riders search for the same trip again and again, and when a bus is late or full, they have to start over.
Found the opportunity from my own commute, led ideation, worked on user adoption, and prototyped with AI tools.
A two-tap routine trip, live alerts with one-tap switching, and a business case for why Transit should build it.
Every weekday, I open Transit to get to campus. I tap "Where to?", search for the same place I searched for yesterday, scroll through the routes, and pick the bus I always take. It's four or five steps for a trip I do every single day. If the bus is late or too full to stop, I back out and do it all over again.
Our brief was to find an Adaptive UI (AUI) opportunity. An AUI learns from what a user already does over and over and helps them do it faster. The brief ruled out recommenders: the system can't suggest something new, it can only enhance something the user already does.
This was a 3-week group project with a team of six, for the course Design of Artificial Intelligence Products.
Each of us brought two AUI ideas to the team. Mine were for Shazam and for a book app. As a group, we filtered out anything that was really a recommender in disguise, which turned out to be most of our first ideas. We started with Uber and the idea of surfacing frequent destinations. That led me to bring up Transit, because I use it daily and had noticed the gap myself.
Routine riders keep searching for the same trip, and they already plan on their phones, so the feature asks for no new behavior.
To check the idea, we tested our wireframes with five peers from another team. Most of them had the same problem. Their main feedback was to look for other moments of friction. That is how we found the second half of the project: what happens when your bus is late or too full to stop.
Our aim was not to change the entire interface, but to consolidate what the app already knows. Delays, crowding and alternative routes are all information the app already has. Our job was to bring it forward at the right moment instead of making riders go looking for it. We also kept Transit's design language and left every original screen reachable, so the adaptive layer feels like part of the app and anyone who prefers the old flow can still use it.
To prototype quickly, my teammates and I gave the same prompt and screenshots to three AI tools: Figma Make, Google Stitch, and Claude Code. I worked on the Figma Make version.
Comparing them showed us a lot about these tools. Figma Make kept drifting away from Transit's original look, Stitch packed in too much information, and Claude Code gave the cleanest result to build from.
Good UX alone rarely convinces a company to build something, so we also made the case for why Transit should invest.
The feature builds on Transit's existing trip planner and real-time data, with a lightweight ranking model rather than a complex AI system.
Transit earns money partly from paid subscriptions. Using about 9M users, 10% paid adoption and a $25 yearly subscription, each extra point of paid adoption adds about $2.25M a year.
These numbers are our own modeled estimates based on public assumptions, not Transit's real figures, and may not be fully accurate.
During our first 5 minute presentation, a feedback round, our professor asked us ‘So what would your AUI actually do’, that’s when we realised we’d spent so long investigating the problem itself, that we hadn’t properly explored the solution. So we then spent the next 1.5 weeks working towards solidifying our solution so it would actually help transit users. That’s when we expanded past just daily routines but also looked at other friction points within the app and realised the late and full bus scenarios were true pain points.
Even though this project taught me a lot about the capability and advantage of AUIs, another important learning was that sometimes the rules of a brief make you shut ideas down too early, and the better move is to find a way to make them work within the rules.
Back to projectsAn adaptive reuse architecture project of a vacant Heinz factory building on Pittsburgh's North Side. Bringing together residential and commercial spaces with the concept of a ribbon.
Reusing an abandoned structure to create a hybrid of two different spaces: a public bathhouse and private residential space.
A key aspect being integration into the environment and community
A site visit, historical research on Allegheny City, conversations with Pittsburgh residents, precedent studies, physical models built to scale, and a full design from plans to renders.
A bathhouse and a traveler's hostel woven together by a continuous ribbon-like wall, growing both horizontally and vertically to mold space.
The studio brief was adaptive reuse where we had to turn an existing building into an "unlikely hybrid" of a bathhouse and a type of housing. My building was the former administrative annex to the main admin building of the Heinz factory on Pittsburgh's North Side.
While the bathhouse was set by the brief, I had the liberty to choose housing styles and I chose a hostel. My research showed me that Pittsburgh's history is much richer than most visitors realize and I wanted to give people an affordable reason to stay close to the parts of the city that are growing right now: the Heinz area, the Strip District, and the bridges.
I started on the site itself. I visited the building, walked the surrounding area, and spoke with Pittsburgh residents about the history of the Heinz factory. I also visited the Senator John Heinz History Center to learn more about the environment my project was placed in.
Alongside that, I researched Allegheny City, the independent city that was annexed by Pittsburgh in 1907 and became what we now call the North Side. Its bridges, including the Three Sisters, slowly pulled it toward downtown, and its name and civic identity faded through renaming and remapping rather than demolition. Two things stood out:
I studied two precedents to supplement my design process:
From this design, I learnt the intricacy of dividing the served and servant spaces seamlessly. It showed me the importance of the division of space without the user feeling the division.
Through this precedent I learnt the importance of utilizing vertical space and extrusions. Instead designing around constraints like compact space.
The key question was how a public bathhouse and a private hostel can share one building without interfering with each other's lives.
My answer was the ribbon. It is one continuous wall that weaves in and out of the existing building, creating space wherever it's needed. Horizontally, it divides the rooms and separates the two programs. Vertically, it folds to form the hostel's sleeping pods. It shows how residential and commercial space can work together in one building.
I also proposed turning the empty lot behind the building into a festival and carnival space for events, giving travelers a reason to come and bringing more activity to the North Side.
The bathhouse includes a main pool, a hot tub, two treatment rooms, locker rooms, a juice bar, a gift shop, two meditation rooms, a light room, and a reading and tea room.
The hostel floors are organized around shared living: four units of sleeping pods, two common kitchens, a dining room, a library and laundry.
I designed the building in Rhino, drew the plans, sections and isometrics in Illustrator and Photoshop, and used Rendair AI to support the rendering process.
The project was well received at final review. The main feedback was to push the ribbon further: how else could it interact with the existing building and with the people using it, so it becomes even more central to the design? If I did it again, I would spend more time on the bathhouse itself and make it more intuitive for visitors to move through.
Back to projectsLearn more about me as a Designer!
So here's what I bring: