Shipped in 2024

AI Sommelier

+10%

Wine sales Increase after 2 months of launch

1st

Google Gemini AI application in Taiwan retail

Collaboration

Design Agency

Role

UX Designer
UX Researcher

Team

UI Designer
UX Researcher
Front-End Engineer
Back-End Engineer
PM

Stakeholder

CTO
Data Team Lead
Dev Team Lead

Platform

Mobile Web App

Duration

2023.09 - 2023.12
(Design Phase)

Business Challenge

Post-Covid wine sales stalled and inventory piled up. Carrefour needed a new way to stimulate sales

User Pain Point

Hundreds of wine bottles with zero guidance. Shoppers don't know where to start — so they just leave

THE RESULT

On-demand AI sommelier turns overwhelmed shoppers into confident buyers

Step-by-step guidance helps shoppers find the right bottle

I transformed sommelier expertise into guided conversation flows with choice-based interactions. This approach made professional wine guidance accessible to beginners while ensuring accurate recommendations, controlled costs, and on-time delivery—all without requiring model fine-tuning

Visual scales and snapshot let shoppers compare wines easily

Simple descriptors like 'micro-bubbles' and 'fruit aroma' paired with numerical scales transformed complex wine attributes into scannable, relatable information.

Food icons and simple description turn wine selection into dinner planning

I combined three elements to make wine expertise accessible: numerical scales (Acidity: 3) for quick comparison, everyday descriptions ('fresh and smooth' instead of 'medium-bodied'), and food pairing icons (beef, cheese, sweets) that relate to real dining experiences

Impact

My design delivered measurable business impact

+0%
+0%

Wine sales Increase after 2 months of launch

0%
0%

Users added wine to cart in ≤3 searches

0st
0st

Google Gemini AI application in Taiwan retail

"AI sommelier can give you the right wine in just a few seconds."

Henry Ting

Digital Technology Officer, Carrefour Taiwan

Full Story

Research

I initiated mixed-method research to quickly capture trends

AI application assessment

I analyzed 14 deployed AI applications to extract workable design and interaction patterns.

I analyzed 14 deployed AI applications to extract workable design and interaction patterns.

Sommelier and seller interviews

We interviewed 2 liquor store owners and 1 Carrefour sommelier to understand wine selection challenges.

We interviewed 2 liquor store owners and 1 Carrefour sommelier to understand wine selection challenges.

Online forum threads analysis

I analyzed wine community forums to uncover beginner barriers and purchasing behaviors efficiently.

I analyzed wine community forums to uncover beginner barriers and purchasing behaviors efficiently.

Insights

We uncovered an opportunity and a pain point

Opportunity: Sommelier's 3-step guidance framework

Expert sommeliers consistently used three questions to guide customers: purpose (gifting, gathering, or personal enjoyment), flavor preferences (taste, origin, grape variety), and budget. This structured approach made wine selection approachable, but required expert availability.

Expert sommeliers consistently used three questions to guide customers: purpose (gifting, gathering, or personal enjoyment), flavor preferences (taste, origin, grape variety), and budget. This structured approach made wine selection approachable, but required expert availability.

User pain point: Knowledge barriers blocked shoppers

Technical wine terminology scared beginners away. Terms like 'dry,' 'oaky,' 'notes of leather,' 'full-bodied with firm tannins,' 'musty,' and 'enology' felt like a foreign language, making wine culture feel exclusive and inaccessible.

Technical wine terminology scared beginners away. Terms like 'dry,' 'oaky,' 'notes of leather,' 'full-bodied with firm tannins,' 'musty,' and 'enology' felt like a foreign language, making wine culture feel exclusive and inaccessible.

Key Feature 1

Expert-informed AI conversation flow

Before

Customers overwhelmed by selection and buy nothing

Carrefour's extensive wine selection intimidated beginners who lacked the knowledge to navigate options. Without accessible expertise, customers left stores without making purchases.

Challenge

How to balance personalization, AI costs, and tight deadlines?

Generating full AI responses and translating them consumed too many tokens, driving up costs and slowing response times. The MVP timeline didn't allow time to extensively test and guarantee AI predictability.

Key Decision 1

Step-by-step guidance helps shoppers find the right bottle

I transformed sommelier expertise into guided conversation flows with choice-based interactions. This approach made professional wine guidance accessible to beginners while ensuring accurate recommendations, controlled costs, and on-time delivery—all without requiring model fine-tuning

Key Decision 2

I designed choice-first interactions to ensure accuracy and delivery timeline

To guarantee accuracy and meet the 4-month deadline, I designed structured choice flows instead of free-form input. Guided selections ensured reliable recommendations without model training, while reducing decision anxiety for wine beginners.

The Design

I shipped expert-guided conversation flow for reliable MVP accuracy

I transformed sommelier expertise into guided conversation flows with choice-based interactions. This approach made professional wine guidance accessible to beginners while ensuring accurate recommendations, controlled costs, and on-time delivery—all without requiring model fine-tuning

Key Feature 2

AI-simplified wine profiles and illustrations

Before

Expert wine language confused beginners, creating a critical barrier

Technical terms like 'pencil lead,' 'sundried rocks,' and 'medium-bodied palate' excluded wine beginners rather than guiding them. This expert language reinforced intimidation, turning potential customers away without purchases.

Challenge

How to maintaining accuracy and credibility with AI?

Apply Accessible Language

Maintain Expert Credibility

Keep Description Accuracy

Feature 1

Visual scales and snapshot let shoppers compare wines easily

Simple descriptors like 'micro-bubbles' and 'fruit aroma' paired with numerical scales transformed complex wine attributes into scannable, relatable information.

Feature 2

Food icons and simple description turn wine selection into dinner planning

I combined three elements to make wine expertise accessible: numerical scales (Acidity: 3) for quick comparison, everyday descriptions ('fresh and smooth' instead of 'medium-bodied'), and food pairing icons (beef, cheese, sweets) that relate to real dining experiences

Key Learnings

Multilingual AI requires cultural adaptation, not just word translation

Direct Chinese translation reduced recommendation accuracy significantly. I learned that multilingual AI needs culturally appropriate concepts and examples to maintain accuracy and relevance.

Direct Chinese translation reduced recommendation accuracy significantly. I learned that multilingual AI needs culturally appropriate concepts and examples to maintain accuracy and relevance.

AI flows can evolve as users became more confident and direct

After launch, user behavior evolved. Initially, beginners needed structured guidance through choices, but as they became familiar with the system, they started entering complete requests upfront. I adapted by allowing direct input alongside guided flows and added 'I don't know' options for uncertain beginners

After launch, user behavior evolved. Initially, beginners needed structured guidance through choices, but as they became familiar with the system, they started entering complete requests upfront. I adapted by allowing direct input alongside guided flows and added 'I don't know' options for uncertain beginners

Next

Next

Next

Next

Smart Warehouse

Smart Warehouse -

I transformed fragmented datas into a unified dashboard.

Smart Warehouse

Next

Smart Warehouse

Previous

Previous