ERP Digital & AI Transformation Expert

The Problem

ERP failures usually come from cross-domain gaps, not missing vocabulary: selection based on demos, implementation without master data, migration without dry runs, integration without transactional consistency, and go-live issues in ledgers, tax, and audit trails. This skill covers the full ERP lifecycle across procurement, supply chain, sales, HR, finance, and materials, plus integration hubs, master data, and extended domains such as production, projects, quality, and equipment. It breaks work into actionable stages: diagnosis → selection/blueprint → investment decision → implementation → data migration → integration → UAT → cutover → operations → AI enablement, and requires verifiable evidence rather than generic advice.

How It Works

• Domain fit first: it confirms industry, scale, organization, and target domain before proposing a solution; retail inventory, discrete-manufacturing BOMs, and group consolidation are treated separately.
• TCO/ROI first: selection or upgrade advice should include 5-year TCO and baseline/expected/pessimistic ROI scenarios, distinguishing financial, compliance, efficiency, and risk-avoidance benefits.
• Traceable standards and compliance: for accounting standards, internal controls, data security, personal information protection, tax/invoice rules, and Xinchuang catalogs, it requires standard numbers, versions, clauses, and retrieval dates, with replacement checks for deprecated references.
• Freshness checks: volatile facts such as tax rates, subsidies, product versions, vendor positioning, and end-of-support dates trigger live verification before conclusions; citations include source and retrieval date.
• Integration and migration closure: it checks integration topology, master-data distribution, migration cleansing, validation, dry runs, rollback design, and whether cross-module data flows have sources, sinks, and consistency.
• Rigor gates: deliverables should pass G1–G12 static gates, DG1–DG6 dynamic gates, and the AI adversarial consensus gate; uncertain items are marked needs review instead of being presented as final.

Boundaries

It is useful for ERP diagnosis, vendor scoring, implementation blueprint review, migration risk checks, integration gap analysis, and AI use-case design. Formal business, financial, tax, or legal decisions still require current regulations, audit evidence, and qualified professionals; for ledger integrity, treasury safety, cross-border data, and regulated industries, complete prototype validation, UAT, parallel runs, and internal-control review first.