✨ Impact Statement

We designed Nosh to solve one of the most common everyday user struggles—what to eat—by leveraging AI to drive healthier decisions, reduce food waste, and lower grocery costs. The product offered a fully personalized experience, automating everything from pantry syncing to grocery list creation. Nosh directly reduced decision fatigue, planning stress, and grocery overspend for busy users.

🥕 My Role

As the sole Product Designer, I led the end-to-end experience for Nosh—from early concept and user flows to wireframes, prototypes, and testing. I collaborated with engineering teams to define API-driven pantry syncing, and worked with LLM experts to craft natural, conversational onboarding and personalized meal logic. My role spanned product thinking, UX strategy, and visual design.

🔍 Introduction & Context

Rising grocery costs and time-starved lifestyles have made meal planning a daily source of stress. Users want tools that are fast, budget-friendly, and personalized. With Nosh, I set out to create a simple, AI-enhanced product that solved this problem end-to-end—what to cook, how to shop, and how to stay healthy on a budget.

🔗 Explore the Figma Prototype →

🎯 The Challenge

Through early discovery and competitive review, I identified four major user frustrations:

• "I never know what to cook"

• "Grocery lists take too long"

• "Healthy meals are hard to plan"

• "Food is getting expensive"

Central design question:

How might we help users make smarter, faster, and more cost-effective meal decisions using AI?

Early concept exploration mapping user input, AI analytics, and grocery integration.

🔬 Insights & Research

From secondary research and market scans:

• 62% of people struggle with meal decisions every day

• 74% prefer automated grocery lists

• Users are looking for AI tools to reduce food waste and planning time

• Competing apps lacked real-time price comparisons and pantry syncing

👤 User Personas

Persona 1: Sarah, Busy Professional	Persona 2: James, Budget-Conscious Parent
Age: 35 Occupation: Marketing Manager Goals: Quick, nutritious meals; minimize grocery trips. Frustrations: Decision fatigue, frequent last-minute takeout, food waste.	Age: 42 Occupation: Teacher Goals: Healthy, affordable meals for family; efficient grocery budgeting. Frustrations: High grocery bills, repetitive meals, difficulty meal planning.

🧩 Mapping User Flows & Use Cases

I defined key use cases and mapped the user journey from frustration to solution:

Defined three core user flows:

• Pantry syncing via grocery account integration.

• Personalized meal planning based on user input.

• Grocery list generation and guided recipe walkthrough.

Comprehensive meal planning journey illustrating AI personalization and user feedback loop.

🎨 Design Process

Understanding the Problem:

I used insights to inform experience strategy, drawing from:

• Competitive Audit of Mealime, Yummly, Paprika

• User frustration mapping

• Content and feature gaps in the market

	Yummly	Mealtime	Paprika
Logo
Strengths	Extensive recipe database; personalized recommendations; supports grocery list generation	Focuses on meal planning based on dietary needs and scheduling; intuitive UI.	Robust recipe organizer; strong pantry and grocery list management; one-time purchase cost rather than subscription.
Weaknesses	Heavy on advertising; UI can feel cluttered; less focus on pantry syncing.	Limited recipe database; sometimes rigid planning tools.	No AI-driven recommendations; limited integration with external services.
Opportunities	Improve real-time price comparison and store integration.	Integrate grocery account syncing; enhance personalization with AI.	Incorporate AI suggestions; connect with grocery APIs for real-time data.

📐 Structuring the Solution

To meet user needs and scale AI:

• Conversational AI onboarding for diet and preferences

• Pantry auto-syncing via grocery APIs

• Modular navigation for pantry, plans, and lists

🧠 I also created an internal AI Logic Library to map inputs to meal outputs, supporting explainability.

AI Data Library and AI Feature Documentation highlighting personalized meal recommendations.

🧭 From Strategy to Structure

Once the core flows and AI architecture were defined, I translated them into low-fidelity wireframes to explore layout, hierarchy, and content clarity. My focus was on creating an experience that felt familiar but deeply personalized by AI logic and pantry data.

View Mid-Fi Prototype

Low-fidelity wireframes illustrating onboarding and pantry syncing

📱 Hi-Fi Design & Key Moments

• Welcome + Value Props: Friendly, colorful introduction screens

• Conversational Onboarding: Chat UI captures goals and preferences

• Grocery Syncing: Account linking for real-time pantry population

• Pantry View: Visual, editable ingredient manager

• Meal Plan Generator: Smart weekly suggestions based on inputs

• Grocery List: Editable, auto-filled list with real-time store pricing

• Recipe Walkthrough: Step-by-step interface with cooking prompts

Welcome + Value Props

Brief, engaging screens demonstrating clear value propositions.

Colorful welcome and onboarding screens.

Conversational AI Onboarding:

User-friendly chat to capture detailed preferences.

Conversational onboarding screens (color).

Grocery Syncing:

Users link accounts for automatic pantry updates.

Grocery account integration screens.

Pantry View/Management:

Automated imports with simple manual adjustments.

Pantry management screens.

AI-Driven Meal Plan Generator:

Custom meal plans generated based on pantry data and preferences.

Meal planning initiation and generated meal screens.

Grocery List Automation:

Dynamic, detailed grocery lists with real-time price comparisons.

Grocery list and refinement screens.

Recipe Walkthrough:

Easy meal modifications via intuitive interactions.

Detailed interaction screens.

✅ Testing & Feedback

I tested mid- and high-fidelity prototypes with 10 participants, uncovering:

• High satisfaction with automatic meal suggestions

• Confusion around pantry syncing → clarified onboarding steps

• Iterated pantry view to better reflect user control and editing

📸 Final Experience Highlights

Delivered an intuitive, AI-powered ecosystem key-flow:

• Conversational onboarding powered by AI

• Smart pantry view synced to receipts and stores

• Weekly meal plan generation

• Grocery list builder with price data

• Cooking assistant with guided steps

• Feedback loop to improve suggestions over time

High-fidelity screens showcasing seamless meal planning and grocery shopping.

📊 Outcomes at a Glance

User feedback:

• 88% found reduced meal planning stress.

• 92% favored automated grocery lists.

• 75% appreciated real-time budget management.

Quantitative projections:

• 40% expected reduction in weekly meal-planning time.

• Estimated 25% monthly grocery savings.

💡 Reflection & Learnings

Key learnings from creating Nosh:

• Conversational UI drove higher engagement and trust

• API integration was key for real-time, personalized value

• Balance between AI automation and user control built confidence

If I could iterate further, I’d explore explainable AI techniques—letting users better understand how meals and prices were chosen.