문서
카테고리
단어
분 읽기
관련 카테고리: 없음
AMWAY global direct-selling context — ABO Field / Insights / D&A · MarTech / CX / Compliance
25-class KG for AMWAY's ABO multi-level + subscription + global multilingual context (~700K edges)
8 common scenarios (S1–S8) + 3 AMWAY-specific scenarios (S9-A · S10-A · S11-A)
In-house ABO · SKU · subscription + 4 external sources (social, weather, FX, regulatory) with cohort_tag segmentation
Bedrock + Neptune + OpenSearch + AgentCore + Cohere — specialized for ABO tree, subscriptions, multilingual operations
Integrated search across Nutrilite, Artistry, Home SKUs + ABOs + external social posts
Subscription state machine + churn signature detection + automated win-back campaigns
Automated review of FTC, Korea Door-to-Door Sales Act, MFDS, and EU direct-selling regulations + multi-country ad guards
A chatbot autonomously invoking tools for ABO Field, CMI, D&A, CX, and Compliance personas
Automated report cards combining Nutrition, Beauty, Home BU revenue with 4 external sources
5 lifestyle personas + KMeans over RFM × category affinity, enriched with social personas
Bayesian attribution across in-house store, ABO direct sale, catalog, and SNS channels
Fusion of 4 sources — social, weather, FX (multi-country), regulatory
A single timeline of in-house store → ABO direct sale → subscription → resignup
Bedrock Guardrails — 4 topics + direct selling + minors + HFF advertising guards
Visualize AMWAY's multi-level ABO tree to a depth of 5 + PV/BV and level performance + Sponsor analysis
A 30-minute demonstration for AMWAY global executives and ABO meetings — Top-5 (S1 · S9-A · S10-A · S2 · S11-A)
5 personas — Brand Marketer / Insights / D&A · MarTech / CRM · LG Members / MD · Channel Sales
3-BU integrated KG with 25 classes — Customer / Product / Transaction / Channel / Operations · External
Eight core scenarios for marketing innovation. Data mix (first-party + 4 external types) is specified for every scenario
PII-masked first-party data + 50K synthetic + 4 external type specifications and cohort_tag separation strategy
Bedrock + Neptune + OpenSearch + AgentCore + Cohere — a single architecture that delivers a working demo in 8 weeks
Integrated semantic search across first-party SKU · Member · Reviews + external social reviews (Olive Young · X)
Per-department autonomous tool-invoking chatbot — Bedrock Sonnet + AgentCore Memory + 12 tools
Automated report cards combining first-party GMV + search trends + weather + competitors
Auto-classified lifestyle personas + RFM × Category affinity KMeans integrated
Omnichannel (Owned mall/Mart/H&B/SNS) Bayesian attribution + search · SNS response combined
First-party GMV × 4 external types (social · weather · economy · competitor) macro analysis
A single member's owned mall → SNS ad → Olive Young → Mart → Repurchase single timeline crossing BUs · channels
Bedrock Guardrails + marketing consent + PII masking + minor cosmetics guard
5 departments — Marketing · Category · Search/Recommendation · CS · Logistics
8 common + 3 Momo-specific (Live attribution · Delivery SLA · Recommendation diversity)
Bedrock + Neptune + OpenSearch + Kinesis (Live) + TimeStream (Delivery SLA)
Unified search across massive SKU + Members + Live reviews + Dcard · Xiaohongshu
24-hour delivery promise vs measured + per-region fulfillment-center load + outdoor weather integration
Recommendation diversity · New-product exposure · Long-tail SKU discovery — exploration-purchase balance monitor
Autonomous tool invocation tailored to each of 5 departments — Marketing · Category · Search-Recommendation · CS · Logistics
Automated reports combining massive-SKU categories + Live + Delivery + External signals
Bayesian attribution across App · Live · TV · Messenger channels
Single timeline — App search → Live viewing → TV home-shopping → Re-purchase
Live viewer → cart → purchase funnel + per-host and per-SKU effect
30-minute live demo for Momo executives, live-operations teams, and logistics teams
Shinkong Mitsukoshi 5-department personas — Marketing, Insights, Digital, VIP, MD
Shinkong Mitsukoshi KG with 25 classes — VIP, foreigner, tenant brand, duty-free, store (~550K edges)
8 common + 3 Mitsukoshi-specific (Foreigner duty-free, Luxury SOV, VIP concierge)
First-party + 4 external signal types — reinforced with FX and tourism signals
Bedrock + Neptune + OpenSearch + AgentCore — multilingual + live FX
Unified search across tenant brand SKUs + VIP/Foreigner + external reviews (Xiaohongshu, Dcard, Instagram)
Counter occupancy analysis across 700+ tenant brands × 19 stores × seasons
Recommendations, pre-sale priority, and personal concierge chatbot for Black/Platinum/Gold members
Autonomous tool invocation tailored to Marketing, Insights, Digital, VIP, and MD personas
Automated reports combining store, floor, and brand GMV with FX, tourism, and competitor data
5 lifestyle personas + RFM × luxury affinity + social personas
Bayesian attribution across store, DM, and social channels for events such as the Anniversary Sale
Fusion of 4 external signals — social, weather, FX (JPY/USD/HKD), and tourism
Single timeline of online → store → duty-free → repeat visit
Bedrock Guardrails 4 topics + duty-free regulations + minors-and-luxury guard
Duty-free recommendation system for foreigners combining passport, nationality, FX, and tourism data
30-minute demo for Mitsukoshi executives, MD, and VIP team meetings
Integrated Marketing · CMI · D&A · OPENPOINT · Manufacturing & Logistics — five departments
Integration of 5 BUs + OPENPOINT + own-manufactured SKUs + cold chain (~900K edges)
8 common + 3 UPI-specific (cross-BU member journey · own-SKU sell-through · cold chain)
5-BU integration + OPENPOINT + own SKUs + cold chain + 4 external sources
Bedrock + Neptune + OpenSearch + AgentCore + TimeStream (cold chain) + Glue (cross-BU ETL)
Unified search across 5-BU SKUs + OPENPOINT members + external reviews
Compare 7ELE / Carrefour turnover of Uni-President Food own SKUs (Uni-Noodle, Mai-Hsiang, etc.) against external channels
President Transnet cold-chain SLA + outdoor temperature + store-order optimization
Autonomous tool invocation across 5 departments — Integrated Marketing · CMI · D&A · OPENPOINT · Manufacturing & Logistics
Combination of 5-BU GMV + own-SKU sell-through + 4 external sources
OPENPOINT unified campaigns + cross-BU-driving channel mix with Bayesian attribution
Fusion of 4 external sources — social · weather · economic · competitor
Cross-BU journey of an OPENPOINT member across own BUs
Single-KG behavior analysis of one OPENPOINT member crossing 7-Eleven → Carrefour → Starbucks → Donut
30-minute demo for UPI group executives and BU operations teams