Skip to main content
Independent concepts

Pantry-first cooking · B2C mobile

Status: Independent concept

MiseLess deciding. More delicious.

A pantry-first cooking concept that turns a handful of ingredients into a good evening, with culinary expertise built into every choice.

Role
Product strategy, UX, visual design, design system and coded prototype
Experiences
Cook’s app + culinary editor · mobile-first
Project type
Self-directed B2C concept · 2026 · discovery through delivery specification
Independent concept in its own visual language · designed and coded · not a shipped client app

Designed as an independent product exploration. Audience needs are hypotheses; research, release plans, and success measures are proposed. The interfaces and working prototypes below are the delivered artifacts. This is not a shipped client application.

01

Frame the product

Dinner starts with what you have, not an endless feed.

A beautiful recipe feed still leaves a person deciding what fits their kitchen. The product hypothesis: start with ingredients and time, then offer a small, explainable set of meals.

THE JOB TO BE DONE
“When it’s nearly dinner and I’m out of ideas, help me make something appealing from what I have without a second shopping trip.”

A focused startup MVP

A consumer cooking membership with a useful free pantry and a curated recipe library.

The MVP scenario assumes a small editorial startup, a curated recipe catalog, and no grocery delivery, automatic pantry recognition, or nutrition prescriptions.

Where I focused

  1. A dinner suggestion from three ingredients
  2. Hands-friendly, one-step-at-a-time cooking
  3. An expert publishing workflow for tested substitutions

What stayed outside the MVP

An infinite social feed, calorie targets, a grocery marketplace, and AI-generated recipes published without review.

The design covers a narrow, complete journey before expanding the feature set.

02

Discovery & product strategy

Understand the moment before dinner.

I began with explicit assumptions and the decisions research would need to inform. This is the study plan, not a set of invented interview findings.

Time-constrained home cooks and culinary editors maintaining a dependable recipe collection. Audience needs are scenario-based assumptions awaiting field research.

HYPOTHESIS 01

Choosing a meal is the difficult first task.

Proposed contextual interviews during dinner planning, followed by a five-evening decision diary.

THE DECISION IT INFORMS

Observe whether people begin with an ingredient, a craving, or a saved recipe before fixing the navigation.

HYPOTHESIS 02

A pantry can become another chore to maintain.

Proposed task tests comparing three quick ingredient picks with a full pantry inventory.

THE DECISION IT INFORMS

Measure time to a relevant idea and whether the setup remains accurate a week later.

HYPOTHESIS 03

Substitutions need trustworthy editorial context.

Proposed culinary-editor review of sample substitutions, including texture, method, and ingredient exclusions.

THE DECISION IT INFORMS

Define what can be suggested automatically and what needs recipe-specific approval.

ASSUMPTIONOBSERVE A REAL TASKTEST AN ALTERNATIVEREVISE THE MVP
03

Onboarding & information architecture

A meal journey, not a content feed.

The core object is a meal that fits a real kitchen. Ingredients, recipe choices, and saved plans stay connected.

  1. 01

    Open the kitchen

    Pick a few ingredients, available time, and preferences. No complete pantry inventory is required.

  2. 02

    Find tonight

    Compare a small set of meals with matches, missing items, and an honest reason for each suggestion.

  3. 03

    Make it yours

    Adjust servings or choose a reviewed swap. Recheck ingredient exclusions after every change.

  4. 04

    Cook

    Move through large, focused steps. Keep ingredients close and preserve the current step if interrupted.

  5. 05

    Keep the good ones

    Save a favorite to the week. Ask whether ingredients should leave the pantry instead of silently deducting them.

Design the return path, too.

No strong match → show a flexible recipe plus clearly missing ingredients. Mid-cook interruption → resume the same step. An unreviewed swap → keep the original recipe and explain the limit.

THREE DURABLE DESTINATIONS
Kitchen / ingredients + tonight
Recipes / curated collection
My week / saved meals + shopping notes
04

From structure to screens

Resolve the behavior before the finish.

These low-fidelity layouts show how the core tasks were organized before the final visual system. Each screen has one main decision and a clear way forward.

Mise•••
01 / Kitchen

What’s on hand?

Peas · Lemon · Pasta
20 minutes available
Find tonight’s dinner
01A few editable picks replace a demanding inventory setup.
Mise•••
02 / Recipe

One good idea

Uses 3 pantry ingredients
Missing: parmesan (optional)
Cook / choose another
02Make fit visible before asking for a commitment.
Mise•••
03 / Cook

One step at a time

Step 2 of 4
Ingredients for this step
Done / back a step
03Large actions and a recoverable position suit a busy kitchen.
05

Interaction & personalization

Make every recommendation earn its place.

01
ALTERNATIVE EXPLORED

A discovery feed with dozens of recipes

A small set with a reason

Explain ingredient matches, time, and any missing items on the recipe itself. Another idea remains one tap away.

The tradeoff: A small set makes catalog quality more visible; a useful no-match state is essential.

02
ALTERNATIVE EXPLORED

A complete pantry before recommendations

Three ingredients to begin

Capture enough context to help now. Let people add items as they cook rather than assigning inventory work upfront.

The tradeoff: Pantry data will be incomplete. Suggestions must distinguish missing items from items merely not recorded.

03
ALTERNATIVE EXPLORED

A generative assistant rewrites the recipe

Reviewed swaps with clear limits

AI can rank approved recipes and surface editorial substitutions. It cannot decide that an ingredient is safe for someone.

The tradeoff: Fewer improvisational answers, with stronger editorial workload and catalog requirements.

AI PRODUCT EXPERIENCE / PROPOSED PRODUCTION CONTRACT

Personalization you can explain.

01 / Context
Chosen ingredients + available time + explicit preferences
02 / Recommendation
Rank the reviewed catalog; explain ingredient fit and show missing items
03 / User control
Edit pantry picks, inspect the full ingredient list, or select another idea
04 / Fallback
No relevant match → browse the curated collection. Ingredient exclusions are hard filters; unavailable filtering blocks personalized suggestions.

The prototype demonstrates these decisions with local rules and sample content. It does not call an AI model.

06

Cook + culinary expert

The expertise behind an effortless dinner.

A calm cooking experience depends on a disciplined editorial workflow. A substitution needs a method, a reason, and a reviewer before it can appear in the consumer app. Try approving or returning the draft.

MISE EDITIONS / RECIPE REVIEWDraft · needs a person

A good swap needs a reason.

Review a recipe-specific substitution with its effect on method and texture. A draft cannot enter a cook’s recommendations.

RECIPE CONTEXTLemon & pea pasta · version 3

Proposed swap: spinach in place of peas.

PROPOSED CHANGEAdd spinach near the finish

Different texture and timing. The original recipe remains available.

Editorial check: ingredient exclusions must run again after a swap. A reviewed recipe is not a personalized safety guarantee.

Illustrative workspace. Try reviewing the draft.

Version before publication

A recipe and its approved substitutions travel together. A retired swap cannot silently remain in a recommendation.

An explanation is part of the content

Editors record how a swap changes texture or timing. The consumer can see that reasoning before making the change.

07

Visual design & scalable framework

Editorial warmth. Kitchen-ready clarity.

An editorial serif for recipe titles; Plus Jakarta Sans for steps and controls; generous line spacing at cooking distance.

TYPE / HIERARCHYSomething
worth making.

20 min · A handful of ingredients

FOUNDATIONS / SEMANTIC COLOR ROLES
Claret / #662B3C
Butter / #F3D77F
Cream / #FFF9EA
Herb / #466849
8PX SPACING RHYTHM / 44PX MINIMUM APP ACTIONS
COMPONENT → PATTERN → EXPERIENCE

Ingredient chip / present + removed

Recipe / matched + missing items + saved

Cooking step / current + completed + resumed

Editorial swap / draft + reviewed + retired

Shared foundations. Typography, spacing, focus treatment, status language, and action behavior connect the consumer and expert experiences. Their layout density follows the task.

Explicit states. Loading retains context. Empty states suggest a next action. Failed saves keep the user’s input and offer a retry. Offline views distinguish saved content from fresh recommendations.

Accessible by design. Selected states use labels and shape as well as color. The app walkthrough supports manual control, keyboard navigation, a global pause, and reduced motion.

08

Design → product → engineering

Make the behavior reviewable, then build it.

I carried the concept through interaction design, visual foundations, reusable React views, and a working prototype. The release specification below describes how a product team could take the next step.

What is built here

Editable pantry picks, ingredient-match explanations, alternate meal ideas, step-by-step cooking, a weekly save state, and an editorial approval flow.

The prototype is intentionally local. It has no account backend, analytics collection, or production content service.

  1. PM: define a cooking completion and separate a saved recipe from an actually cooked meal.
  2. Engineering: persist the current recipe version and step; retain local progress offline; recheck exclusions after any swap.
  3. Editorial: version recipes and substitutions together; require a named reviewer before publication.
  4. QA: cover no matches, missing pantry items, text enlargement, offline steps, a failed save, and a retired substitution.

AI in the design workflow

AI-assisted exploration helped compare interaction directions and draft the coded prototype. The review focused on the resulting screens: clear hierarchy, consistent state changes, keyboard operation, responsive layouts, and honest sample content. Generated alternatives do not stand in for user research.

09

Engagement, retention & optimization

Define a useful return before measuring one.

No conversion lift, retention gain, or research result is claimed. These are proposed experiments with defined cohorts and guardrails, ready for a PM, engineer, and researcher to refine together.

MEASUREMENT PLAN

Useful discovery

New cooks starting guided cooking within their first two visits ÷ eligible new cooks with two visits or a seven-day window.

Events to specify
pantry_started → recipe_viewed → cooking_started
Experiment
Three-ingredient setup vs. browsing a curated collection
Guardrail
Track no-match rate and ingredient-exclusion failures; irrelevant starts do not count as a win.
MEASUREMENT PLAN

A repeat cooking habit

Activated cooks completing guided cooking on two distinct days in days 8–21 ÷ activated cooks eligible for that full window.

Events to specify
cooking_completed · meal_saved · recipe_returned
Experiment
A saved weekly shelf vs. a reminder-only return route
Guardrail
Watch notification opt-outs and self-reported ingredient waste; use completion feedback, not screen time.

A small release, with a learning agenda.

01 / VALIDATE THE TASKRun the discovery plan and moderated prototype tasks. Revise the onboarding and recovery paths before building a live service.

02 / PILOT THE WHOLE LOOPRecruit a small, consented pilot. Include the editorial workflow, content operations, and failures, not just the consumer happy path.

03 / COMPARE RESPONSIBLYAgree on cohorts, experiment duration, and a sample-size plan before randomizing. Keep the guardrails alongside the primary metric.

DELIVERED / REFLECTION

A complete product argument, made tangible.

The most useful interface begins before the recipe: it reduces the effort of choosing. The open question is whether a small, relevant set continues to feel helpful as tastes change and pantry data ages.

Delivered: product framing, a research plan, journey and architecture, wireframes, interaction decisions, visual foundations, expert tools, a handoff specification, and a coded mobile prototype.

10 / THE FINISHED EXPERIENCE

Meet Mise.Make something of tonight.

THE FINISHED MOBILE EXPERIENCE

From “what’s here?”
to “that was good.”

Watch the product journey unfold, or take control of the app. Choosing a screen or using a phone control pauses the walkthrough.

01 / 04

Interactive concept · Sample content · Changes stay in this page. Recommendations are scripted; no AI service, account, payment, or message is connected. Photography from Unsplash, used under the Unsplash License.

AJ

What’s in your kitchen?

Tap a few things. Dinner ideas follow.

Ingredients on hand
A bowl of pasta with peas, lemon and mint
Tonight could beLemon & pea pastaUses 3 of yours · 20 min
01 / Your kitchen

A few ingredient picks create a useful starting point.