Problem Statement: Busy professionals feel drained by the meal planning and grocery shopping process, making it difficult for them to eat a nutritious meal at home.
Persona: Maya, the "Healthy Habit-Seeker." Maya is a 35-year-old busy professional who wants to feel in control of her health and save money, but the mental load of finding recipes and shopping drains her limited energy.
Key Features: Insta-Prep is an interactive, multi-screen web application designed to eliminate meal-prep fatigue. The app features a scrolling feed of simple, 3-ingredient healthy recipes. With a single "Add to Cart" click, the app automatically compiles the necessary ingredients into a digital grocery list array.
Testing & Iterations:
During our testing phase, we relied on an AI Focus Group and human user testing applying "The Mom Test" to gather honest, constructive feedback. Our AI Focus Group—prompted to act as a rigorous thesis review council—identified a critical flaw: users could add ingredients to a cart, but could not remove pantry staples they already owned (like olive oil or salt). Because Maya's secondary goal is to save money, forcing her to overbuy ingredients violated our core value proposition.
Additionally, our human tester revealed significant layout friction. During silent observation, the user successfully added an item but could not figure out how to view their cart because our navigation menu was pinned to the bottom of the screen. Based on this feedback, we iterated on our prototype by rewriting our CSS to pin the navigation bar to the top right of the screen (matching the user's mental model of standard grocery apps). We also prompted our AI coding assistant to help us write a removeFromCart() JavaScript function, giving the user full editing control over their final list.
Final Reflection: This design thinking journey taught me that jumping straight to a technical solution is a trap. If I had simply built my first idea, I would have created an app that frustrated users and cost them extra money. Using AI as a coding assistant allowed me to build the HTML and JavaScript structure incredibly fast, but it was the human-centered approach—specifically forcing the AI to be critical and watching real users struggle with my navigation—that actually made the application valuable and usable.
Prototype Link: [Insert Web Studio Link Here]