AI-POWERED SCHEDULING
A decision-support tool that eliminates choice paralysis.
The average person spends 15 minutes a day deciding what to eat, often leading to "Decision Fatigue." Appetite was built to eliminate the paradox of choice. The challenge: build a discovery tool that aggregates multi-platform data into a 3-step decision flow, helping users go from "hungry" to "ordered" in under 60 seconds.
Timeline
From explorations to final designs in 7 weeks while working with multiple projects at the same time
My Contributions
As the Lead Product Designer and Strategist, I defined the core "Choice-Reduction" algorithm, designed the end-to-end mobile experience, and conducted high-tempo A/B testing on the filtering logic. I optimized the handoff flow to external delivery APIs, resulting in a significant lift in successful order completions
50% Reduction in average decision time per user.
100% Task Completion achieved during high-fidelity usability testing.
66% Projected Increase in signup conversion via "Zero-Type" onboarding.
3-Minute Save per user per session compared to standard delivery aggregators.
Planning the app structure
Once the high-level structure was defined, I moved into low-fidelity wireframing to map out the core screen layouts. The goal here was to stress-test the user flow without the distraction of visual aesthetics, ensuring the "path to food" remained as short as possible.
The primary focus during this stage was the Onboarding Experience and the Decision Engine—Appetite’s core tool. I designed the wireframes to prioritize a "Choice-Reduction" layout, moving away from the standard long-list format found in delivery apps. By focusing on how users filter cravings through a simplified interface, I laid a scalable foundation that allowed for the later introduction of multi-vendor API integrations.
This category details the step-by-step approach taken during the project, including research, planning, design, development, testing, and optimization phases.
Solving the Paradox of Choice (Discovery)
Referencing Evil by Design (Sloth), I identified that "More is Less." Users feel overwhelmed when presented with 100+ options. I pivoted the product mission from "Discovery" to "Selection." I conducted 10 interviews to map the "Hunger Journey" and found that users value Speed of Decision over Volume of Options.
The "Decision Engine" Architecture (Planning)
I mapped the app structure to act as a funnel. To tackle the complexity of multi-vendor data (UberEats, DoorDash, etc.), I devised a role-based access system where the user only sees the most relevant "Choice Molecules" based on their real-time geofence and dietary history.
Systematic Design (Atomic Methodology)
I referenced Brad Frost’s Atomic Design: To ensure the app remained consistent across vendors, I built a standardized Merchant Organism.
Atoms: Standardized labels for price, amount, and mood type.
Molecules: Normalized restaurant cards that make scanning 10+ options per second effortless. This system ensures that no matter where the data comes from, the UI remains calm and predictable (referencing Don't Make Me Think).
Removing Friction (Interaction Design)
One major sticking point was the "Address Entry." I implemented a geofencing API and a "Quick-Select" pattern that removed the need to type on mobile. By using Progressive Disclosure, we only ask for specific preferences (Vegan, Halal) after the user has engaged with the primary feed, significantly lowering the barrier to entry.
Validating the MVP (Testing)
Even though the app is not live, I conducted usability tests on the prototype. I discovered that users found the initial "Filter" screen confusing. I fixed this by rethinking the tool around a "Smart Craving" Swipe Editor, where users can see 3 curated options at once—a UI pattern that later increased selection speed by 10x in testing.
After the initial prototype was developed, I moved into a cycle of rapid testing and refinement. This phase was critical for addressing user pain points and aligning the interface with the actual mental models of our users.
Increased Efficiency
The initial version of Appetite’s "Choice Screen" featured a traditional list of restaurants. Feedback from usability testing revealed that users were still experiencing "Selection Fatigue," often taking more than 2 minutes to make a choice even with filters applied.
Redesigning the Selection Logic
I addressed this by rethinking the interaction model. I moved away from the list view toward a Single-Card Recommendation pattern. By presenting one "Best Match" at a time with clear, high-contrast metadata
Growing User Base
We achieved a 3-minute reduction in selection time compared to the initial wireframes.