Wine Wise Companion
artificial intelligence (AI) wine scanner, taste profiles and cellar
Outcome
One photo to a confident wine choice.
- ✓Pick a bottle without expert knowledge
- ✓Know whether the price is fair
- ✓Plan wine for a group in minutes
- The challenge
- Most wine drinkers stand in an aisle or read a wine list with no idea which bottle they'll like or whether the price is fair, and expert apps assume expert knowledge.
- How I led it
- Product owner and designer. I designed the scan-to-recommendation flow, taste profiles, party picker, social features and subscription tiers, and directed AI-assisted development.
- What I delivered
- Wine Wise turns one photo into a recommendation: AI identifies wines on a label, shelf or list, matches them to the user's taste profile and checks current prices, then helps plan party picks and track a cellar.
Deep dive
Open any section for detailHow it worksFrom photo to the right bottle
- 01
Snap
Bottle, shelf or wine list
- 02
Identify
AI recognizes each wine
- 03
Match
Ranked to your taste profile
- 04
Price
Live price check
- 05
Enjoy
Party picks, cellar, diary
Challenge and solution, side by side3 examples
- ChallengeA wall of unfamiliar bottles.SolutionScan the shelf. AI identifies the wines and ranks them against the user's taste profile.
- ChallengeIs this price fair?SolutionLive price suggestions from a real-time search service.
- ChallengeChoosing wine for a party.SolutionThe party picker recommends wines for the group and shares the picks by link.
Product focus and qualityFocus, validation approach and key features
Product focus
One photo to a confident choice, in plain language, tuned to the person's own taste.
Quality & validation
Uploaded photos are AI-moderated. Real-time pricing comes from a live search service rather than the model's memory. Terms of service and an intellectual-property notice are in the app.
Key features
- ·Scan a bottle, shelf or wine list from a photo, plus barcode lookup
- ·Taste profiles built manually or by AI from liked wines
- ·Real-time price suggestions
- ·Party picker with shareable picks for a gathering
- ·Cellar inventory with CSV/Excel and CellarTracker import
- ·Availability alerts, favorites and a pairings diary
- ·Friends, QR connect and critic profiles
- ·AI help chat and subscription tiers
Why it matters
Shows I can make expert knowledge easy to use, from a single photo.
Scope, notable components and compliance
Measurable scope
- ·7 AI and data services (scan, barcode, pricing, party picks, alerts, help, moderation)
- ·13 versioned database changes
Notable & novel components
- ·Greg's Scan & Analyze (marked patent pending in the app): bottle, menu, food and food + drink scanning
- ·Multi-bottle recognition from a single shelf or list photo
- ·AI-built taste profiles from liked wines
Compliance work
- ·AI moderation of uploaded images
- ·Terms of service and IP notice
Inside the app15 screens
Tech stack and skills
Tech stack
React + TypeScriptViteTailwind / shadcnLovable Cloud (Postgres, auth, storage, serverless functions)Lovable AI Gateway (vision)Perplexity (live pricing)
Skills demonstrated
Vision AIRecommendation designConsumer user experience (UX)Data import
▤ README.md›
# Wine Wise Companion > artificial intelligence (AI) wine scanner, taste profiles and cellar "Shop like a sommelier. No degree required." Snap a bottle, a store shelf or a wine list and AI identifies the wines, matches them to your taste and finds current prices. It also plans party picks and tracks your cellar. ## Outcome **One photo to a confident wine choice.** - Pick a bottle without expert knowledge - Know whether the price is fair - Plan wine for a group in minutes ## Problem Most wine drinkers stand in an aisle or read a wine list with no idea which bottle they'll like or whether the price is fair, and expert apps assume expert knowledge. ## Solution Wine Wise turns one photo into a recommendation: AI identifies wines on a label, shelf or list, matches them to the user's taste profile and checks current prices, then helps plan party picks and track a cellar. | Problem | Solution | | --- | --- | | A wall of unfamiliar bottles. | Scan the shelf. AI identifies the wines and ranks them against the user's taste profile. | | Is this price fair? | Live price suggestions from a real-time search service. | | Choosing wine for a party. | The party picker recommends wines for the group and shares the picks by link. | ## Features - Scan a bottle, shelf or wine list from a photo, plus barcode lookup - Taste profiles built manually or by AI from liked wines - Real-time price suggestions - Party picker with shareable picks for a gathering - Cellar inventory with comma-separated values (CSV)/Excel and CellarTracker import - Availability alerts, favorites and a pairings diary - Friends, Quick Response (QR) connect and critic profiles - AI help chat and subscription tiers ## Tech stack - React + TypeScript - Vite - Tailwind / shadcn - Lovable Cloud (Postgres, auth, storage, serverless functions) - Lovable AI Gateway (vision) - Perplexity (live pricing) ## Notable components - Greg's Scan & Analyze (marked patent pending in the app): bottle, menu, food and food + drink scanning - Multi-bottle recognition from a single shelf or list photo - AI-built taste profiles from liked wines ## Compliance - AI moderation of uploaded images - Terms of service and IP notice ## Scope - 7 AI and data services (scan, barcode, pricing, party picks, alerts, help, moderation) - 13 versioned database changes ## Links - Live: https://gregswinewise.lovable.app ## Author Candace Hagood · Product & Quality Leader · AI-Assisted Engineering