6 Sub-dimensions · Click to expand L2 detailClic para expandir detalle L2
🏗️
› L2N2
L2 · Supply Chain Data Architecture & Infrastructure
Data lakes, warehouses & lakehouse architecture, real-time streaming (Kafka/IoT), API integration strategy, cloud vs. on-premise TCO, and data security & privacy compliance.
The foundational layer of the intelligent supply chain — the architecture that captures, stores, integrates, and secures data from ERP, WMS, TMS, IoT, and external sources, making it available for analytics and AI models at the right latency.
L3 Sub-componentsSubcomponentes L3 5 items · click to explore elementos · clic para explorar
01
Data lakes, data warehouses & lakehouse architecture for supply chain
Lakehouse as the dominant 2025 paradigm (Delta Lake/Databricks), data freshness SLAs by use case, and the unified pipeline that reduced forecast MAPE from 22% to 16% by enabling real-time POS data.
→
02
Real-time data streaming: Kafka, event-driven supply chains & IoT integration
Apache Kafka as the event streaming standard, cold chain temperature alert in <2 minutes vs. 4-hour batch delay, and the 67% reduction in irreversible temperature excursions.
→
03
API integration strategy: connecting ERP, WMS, TMS & external systems
REST vs. EDI vs. event-driven integration, the API catalog as the prerequisite of reuse, and the OTIF improvement from ASN latency reduction from 24h to 5 minutes.
→
04
Cloud vs. on-premise for supply chain data: migration & TCO analysis
Snowflake vs. Oracle Exadata TCO ($180K vs. $1.08M/year), cloud cost management requirements, and 4-week deployment speed as the key cloud advantage.
→
05
Data security, privacy & compliance in supply chain data platforms
GDPR/LFPDPPP requirements, RBAC from day 1, dynamic data masking for personal data in dev/test, and the audit log as the compliance evidence.
→
📦
› L2N2
L2 · Master Data Management & Data Governance
Product master data (SKU hierarchy & lifecycle), customer & supplier master (golden records & deduplication), data governance framework, GS1 standards & GTIN, and data catalog & business glossary.
The data foundation of supply chain analytics — without accurate, complete, and consistent master data, every forecast model, every dashboard, and every optimization algorithm produces systematically incorrect results.
L3 Sub-componentsSubcomponentes L3 5 items · click to explore elementos · clic para explorar
01
Product master data: SKU hierarchy, attributes & lifecycle management
The 4 most costly master data problems and the $180K/year freight overpayment from incorrect weights.
→
02
Customer & supplier master data: golden records & deduplication
The 15% duplicate rate in supplier master data and its impact on spend analytics, and the golden record as the canonical source of truth.
→
03
Data governance framework: ownership, stewardship & quality rules
The 4 data roles (Owner/Steward/Consumer/Engineer), the monthly Data Quality Scorecard, and the 75% reduction in data quality operational costs.
→
04
GS1 standards, GTIN & supply chain data interoperability
GTIN/GLN/SSCC/EPCIS as the global supply chain language and the 38% sales lift from fixing missing product images via GDSN.
→
05
Data catalog & business glossary: making supply chain data discoverable
The business glossary as the lowest-cost, highest-impact investment in supply chain analytics, and reducing analytics prep time from 35% to 8%.
→
🤖
› L2N2
L2 · ML & Predictive Analytics in Supply Chain
Predictive maintenance (IoT + ML), route optimization ML, supplier risk scoring, demand forecasting with ML (external signals & ensemble), and prescriptive analytics & optimization.
The intelligence layer of the supply chain — the ML models that predict failures before they occur, optimize routes in real time, score supplier risk before the disruption happens, and recommend the exact inventory level that minimizes both cost and stockout risk simultaneously.
L3 Sub-componentsSubcomponentes L3 5 items · click to explore elementos · clic para explorar
01
Predictive maintenance: ML models for asset failure prediction
Anomaly detection for bearing degradation 3–4 weeks before failure, the 2.8× first-year ROI of PdM implementation.
→
02
Route optimization ML: dynamic routing, time windows & multi-depot
The 23% distance reduction and 29pp OTIF improvement from ML-optimized routes, and ETA prediction accuracy as the customer experience metric.
→
03
Supplier risk scoring: ML models for supply chain risk intelligence
5 risk signal categories, the 65% predictive accuracy target, and $1.8M USD in avoided production stoppages.
→
04
Demand forecasting with ML: external signals, NLP & ensemble methods
The statistical+ML ensemble as the 2025 best practice, SHAP values for explainability, and the 32% MAPE improvement.
→
05
Prescriptive analytics & optimization: from prediction to action
Safety stock optimization that simultaneously reduced capital by $13.4M MXN and stockout rate by 56%, and the 70% adoption rate target.
→
📊
› L2N2
L2 · Data Visualization, Dashboards & Decision Support
Supply chain control towers, KPI dashboard design principles, advanced analytics visualizations (network maps/heatmaps/Sankey), self-service analytics, and embedded analytics & AI copilots.
The decision support layer — the control towers, dashboards, and AI copilots that convert supply chain data into the right visual representation at the right time for the right decision-maker.
L3 Sub-componentsSubcomponentes L3 5 items · click to explore elementos · clic para explorar
01
Supply chain control towers: real-time visibility dashboards
Exception-based alerting vs. reporting dashboards, the 30-minute alert response time target, and the 42% WISMO reduction from proactive delay notifications.
→
02
KPI dashboard design: principles, hierarchy & storytelling with data
The 3-level dashboard pyramid (executive/managerial/operational), the 10–15 KPI maximum, and time-to-insight as the usability metric.
→
03
Advanced analytics for supply chain: network maps, heatmaps & Sankey diagrams
Network maps as the highest-impact one-time analysis in logistics network design and the 3 insights in 1 hour that 3 years of tabular reports never surfaced.
→
04
Self-service analytics: enabling business users with data without IT dependency
The 4-level self-service maturity model, data catalog + business glossary as prerequisites, and the 3-week to 2-day analytics backlog reduction.
→
05
Embedded analytics & AI copilots in supply chain applications
SAP Joule and O9 AI copilot for S&OP narrative generation, the 62% S&OP preparation time reduction, and the human-in-the-loop validation requirement.
→
🔧
› L2N2
L2 · AI Tools & Platforms for Supply Chain
Build vs. buy decision framework, MLOps for model versioning & retraining, NLP for contract analysis & risk signals, computer vision for quality inspection, and responsible AI (bias, explainability & governance).
The AI capability layer — the platforms, operations practices, and governance frameworks that determine whether supply chain AI models are trustworthy, maintainable, and actually adopted by the operational team.
L3 Sub-componentsSubcomponentes L3 5 items · click to explore elementos · clic para explorar
01
AI platform selection: build vs. buy for supply chain AI applications
The decision framework, the proprietary data advantage that tips the balance toward build, and the 12% vs. 16% MAPE advantage of locally-trained models.
→
02
MLOps for supply chain: model versioning, monitoring & retraining pipelines
Model drift as the silent killer (22% MAPE after 6 months unmonitored), MLflow for versioning, and the 7-day detection target with automated monitoring.
→
03
Natural language processing (NLP) for supply chain: contract analysis & risk signals
Contract analysis at 5 minutes vs. 8 hours, 142 force majeure gaps found in 2 days, and the 4,000-hour productivity saving across 500 contracts.
→
04
Computer vision in supply chain: quality inspection, OCR & warehouse automation
99.2% defect detection vs. 72% human inspection, the 2.1× first-year ROI, and YOLO v8 as the real-time detection standard.
→
05
Responsible AI in supply chain: bias, explainability & governance
The 24% size bias in supplier scoring, SHAP values as the explainability standard, and the adoption rate jump from 38% to 74% after the bias correction.
→
🔄
› L2N2
L2 · Analytics-Driven Operations & Continuous Improvement
Data-driven culture (gut-feel to evidence-based decisions), A/B testing & experimentation, Six Sigma & DMAIC with analytics, digital twin for scenario planning, and analytics ROI measurement.
The organizational layer that converts data capabilities into operational outcomes — the culture, experimentation practices, improvement methodologies, and ROI frameworks that make analytics a sustained source of competitive advantage in supply chain.
L3 Sub-componentsSubcomponentes L3 5 items · click to explore elementos · clic para explorar
01
Data-driven culture: from gut-feel to evidence-based decisions in supply chain
The 5-level maturity model, data literacy as the prerequisite, and the 68% data-driven decision rate achieved in 12 months with the data champion program.
→
02
A/B testing & experimentation in supply chain: validating operational changes
The causal proof advantage of A/B testing over historical analysis, the 4–8 week minimum experiment duration, and the $42K MXN test that saved $0.80/order × all CDMX volume.
→
03
Six Sigma & DMAIC with data analytics: process improvement in the digital age
DMAIC in 3 weeks vs. 6 months with lakehouse data, Random Forest as the root cause analysis tool, and the 31% picking time reduction.
→
04
Digital twin for supply chain simulation & scenario planning
The 4 major use cases, the 30-minute simulation time target, and the 10% calibration accuracy requirement for decision-grade predictions.
→
05
Analytics ROI measurement: quantifying the value of data & AI in supply chain
The 4-dimension ROI framework, the 5.4× total program ROI, and how to present analytics as a capital investment to the CFO before budget season.
→