Shinkong Mitsukoshi Luxury Retail PoC
Published 2026-05-14Updated 2026-07-024 min read
"A department store innovation PoC that ties together VIP members, foreign members, luxury tenant brands, duty-free, and store data with external trend, FX, and tourism signals through Ontology + Agentic AI"
VIP membership across 19 Shinkong Mitsukoshi stores, 700+ luxury tenant brands, foreign tourist behavior (Japan, Hong Kong, Southeast Asia), and duty-free data integrated into a single KG. Uses public information only.
🚀 Interactive Demo
👉 Open demo page (new window)
Switch between 11 scenarios from the sidebar to experience the persona switcher, foreign tourist analysis, luxury SOV, and VIP concierge UIs hands-on.
1. PoC One-Line Summary
| Axis | Description |
|---|---|
| Target audience | 5 departments: Marketing, Customer Insights, Digital, VIP Membership, MD |
| Subject | VIP members, foreigners, 700+ tenant brands, duty-free, stores + 4 external signal types |
| Method | Ontology (KG with 25 classes) + Agentic AI (Bedrock + AgentCore) |
| Demonstration | Live execution of 11 scenarios + 5-department persona switcher |
| Differentiation | Foreign duty-free recommendation + luxury SOV + VIP concierge — 3 specialized capabilities |
2. Business Structure (Public Information)
| Axis | Description |
|---|---|
| Stores | 19 (Taipei Xinyi/A11/A8/A9/A4 + Taichung/Tainan/Kaohsiung/Taoyuan/Chiayi) |
| Tenant brands | 700+ (LV, Hermes, Chanel, Gucci, Dior, Prada, Tiffany, etc.) |
| Membership | skm-member (VIP/Black/Platinum/Gold/Standard) |
| Duty-free | Foreigner TaxRefund + duty-free shop operations |
| Key seasons | Anniversary Sale (週年慶, major fall sale), Chinese New Year (春節), SS/FW |
3. 5-Department Personas
| Code | Department | KPI |
|---|---|---|
| P1 | Marketing & Campaigns | Anniversary Sale ROAS · Foreigner acquisition · New product launches |
| P2 | Customer Insights | VIP LTV · Foreigner share · SOV |
| P3 | Data & Digital | Model accuracy · Duty-free data quality · Mobile SLA |
| P4 | VIP & Membership | Tier upgrades · Lounge utilization · Concierge NPS |
| P5 | MD & Store Operations | Store and floor revenue · Tenant brand turnover · Counter occupancy |
4. 11 Scenarios (One-Line Summary)
| # | Scenario | Data Mix |
|---|---|---|
| S1 | Natural-language semantic search | Tenant brand SKU + members + external reviews (Xiaohongshu, Dcard) |
| S2 | 5-department persona chatbot | Autonomous invocation of all tools |
| S3 | Category / BU insight cards | First-party GMV + FX + tourism + competitors |
| S4 | Lifestyle persona matching + clustering | RFM + luxury affinity + social |
| S5 | Omnichannel campaign ROAS simulator | Anniversary Sale, DM, SNS, influencer channel mix |
| S6 | External signal fusion (4 types) | First-party GMV × 4 external signals (including tourism metrics) |
| S7 | Omnichannel member journey | First-party online → store → duty-free → repeat visit |
| S8 | Marketing consent / PII guardrails | Bedrock Guardrails + duty-free regulations |
| S9-M ⭐ | Foreign tourist behavior analysis + duty-free recommendation | Passport, nationality, FX, tourism |
| S10-M ⭐ | Luxury brand SOV + counter occupancy | 700+ tenant brands × stores × seasons |
| S11-M ⭐ | VIP tier lounge concierge | Black/Platinum/Gold recommendations and pre-sale |
5. AWS Stack (Common)
Same as LG H&H + multilingual (Traditional Chinese, Japanese, English), live FX added, ~550K edges (enriched tenant brand catalog).
6. Top-5 Demo Flow (30 minutes)
| Order | Scenario | Demo Highlight | Time |
|---|---|---|---|
| 1 | S1 Semantic search | "Female in her 30s + LV/Chanel + Anniversary Sale 50% off" | 4 min |
| 2 | S9-M ⭐ Foreigner duty-free | "Japanese tourist in her 30s → duty-free recommendation + FX-aware" | 5 min |
| 3 | S10-M ⭐ Luxury SOV | "Hermes vs Chanel counter occupancy by store" | 4 min |
| 4 | S11-M ⭐ VIP concierge | "Pre-sale recommendation for a Black-tier member" | 5 min |
| 5 | S2 Persona chatbot | "Anniversary Sale ROAS" P1 vs P5 comparison | 4 min |
Details: 99. Demo Storytelling