Krave

Feed the feeling

Team: Debbie Chen, Sonya Pholsiri, Kaitlyn To, Alyssa Alano
School: The University of Texas at Austin
Event: FigBuild Hackathon, March 2026

Inspiration

At the beginning of Figbuild, our team was actually heading in a completely different direction. During a break, Alyssa got hungry and started scrolling through DoorDash trying to decide what to eat, but she couldn’t figure out what she wanted.

Frustrated, she said something along the lines of,

“What if an app could just tell you what you’re craving?”

That simple idea immediately clicked with the rest of us, and from there the concept of Krave took off.

We began thinking about how people often struggle to decide what to eat—not because they’re indecisive, but because they don’t fully understand the signals their body is giving them. That moment became the spark that shaped our entire project.

What It Does

Krave helps people make food decisions by translating internal signals like mood, energy level, and cravings into personalized food recommendations. Inspired by the concept of alliesthesia, the idea that our body's internal state changes what foods feel satisfying, Krave helps users better understand what their body actually wants to eat.

Personalized Recommendations

The home screen shows a live breakdown of the user's current emotional and physiological state, where each emotion is represented as a bubble. The bigger the bubble, the stronger the feeling. Krave uses this to generate food recommendations matched to the user's strongest signal, with details on nearby restaurants or recipes. Over time, Krave also builds personal Spreads, collections of foods grouped by emotional context that update automatically as the user logs more meals.

Social and Group Decisions

Users can see what their friends are currently craving in real time, and use the Mood Map to see which emotions their friend group is feeling most strongly. When eating with others, the Krave Krew feature merges everyone's emotional states into one shared breakdown and generates a single food recommendation that accounts for the whole group.

Craving Reflection

Krave automatically logs meals daily and tracks the user's emotional state before and after eating, so users can see whether a food actually resolved the craving that triggered it. A monthly summary and shareable cards then surface patterns across the user's full history, showing connections like which foods consistently help and which ones they reach for out of habit.

Safeguards

Dietary Safety

During onboarding, Krave collects the user's allergies, intolerances, and dietary restrictions, including common allergens, vegetarian, vegan, halal, and kosher needs. These are treated as hard filters, not preferences. No food containing a flagged ingredient will ever appear in a recommendation, a Spread, or a Krave Krew result. If a user joins a group session, their restrictions apply to the group output as well. No one in a Krew will be recommended something that conflicts with another member's dietary profile.

Emotional Data Privacy

Emotional state data is personal and sensitive. Krave does not sell, share, or use this data for advertising. Emotions and cravings are only visible to friends a user has chosen to connect with in the app. Users can remove connections or opt out of the social features entirely and use Krave in a fully private mode at any time.

No Pressure Design

Krave suggests, never prescribes. The app surfaces what your body might be signaling, but the choice of what to eat is always entirely the user's. There are no alerts and no judgment for whether a user followed a recommendation or ignored it entirely.

How We Built It

We built Krave entirely in Figma over the course of the hackathon. Early ideation and user flow mapping took place in FigJam, where the app’s features and logic were worked out before any UI design began. Visuals were created directly in Figma, including the Nom characters and the blob based interface elements, built from scratch using Figma's vector tools. The full prototype was then connected using Figma’s native prototyping features.

We used Figma Make for early visualization, helping us understand how the app might function and feel before the UI was fully designed. Throughout the design process, Claude Opus 4.6 played a key role in clarifying our thinking, especially around information hierarchy and deciding which features to prioritize and how they connected to each other. We also used YouTube tutorials and a range of AI tools including ChatGPT, Gemini, and Copilot during the early ideation phase to explore potential concepts and directions before committing to a direction.

Challenges We Ran Into

Attempting to quantify something as subjective as cravings into a structured system was a significant hurdle. Cravings are influenced by many different variables from emotions to environment. Building a model that was easy to navigate yet sophisticated enough to account for these diverse variables was challenging.

One of our biggest challenges was the learning curve. This was Alyssa’s first time using Figma, so she had to quickly learn the platform while we were building the project. It had also been a while since Kaitlyn last used Figma, so she had to relearn many of the tools as well.

Time management also became a challenge. The timeline was tighter than we expected, and the daylight savings time change made things even more chaotic than anticipated. Because of this, we ended up staying up very late to finish everything in time.

Accomplishments That We’re Proud Of

We’re proud that we were able to create a concept that balances imagination and practicality. Krave explores a creative idea—translating cravings into food recommendations—while still grounding it in a realistic product experience.

We’re also proud of how much our team improved our Figma skills in such a short amount of time. Most importantly, we had a lot of fun working together. True to our app’s theme, we shared many cravings (especially matchas!) together throughout the process. The challenge brought us closer as a team and made the experience even more rewarding.

What We Learned

Through this project, Alyssa learned how to use Figma Design, and all of us learned how to work with Figma Make. Sonya learned how to tailor her prompts in Figma Make, and work between Design and Make files. Debbie also learned how to create interactive prototypes in Figma and design vector graphics directly within the editor using Figma’s built-in vector tools.

Beyond technical skills, we also learned how to collaborate efficiently, divide responsibilities, and adapt quickly when challenges came up. We also learned how to balance our desire for innovative, fantastical elements of an app with realistic product requirements given our tight deadlines and skill levels. Though we did not get to design every feature we originally dreamed up, we were happy to have a finished design that captured the spirit of our original vision.

What’s Next for Krave

The version of Krave we built this weekend is just the beginning. Next, we plan to flesh out the physiological needs side of the app, which is still largely untapped compared to the emotional features we focused on first.

We're particularly interested in exploring Krave's applications for people with specific health needs. For patients recovering from illness or surgery, food plays a direct role in healing, and Krave could help surface what the body actually needs nutritionally during recovery rather than just what sounds good. We'd also love to expand Krave's support for hormonal cycles, both for menstrual health and pregnancy, where cravings are deeply physiological and often misunderstood. These are areas where understanding your body's signals isn't just about comfort, it's about care.

At its core, Krave is about helping people understand what their body is telling them. We want to keep pushing that further, across more users, more needs, and more moments when listening to your body matters most <3