---
id: "action-fix-data-infrastructure"
type: "action-item"
source_timestamps: ["§ 1. Create and maintain high-quality data."]
tags: ["data-engineering", "project-sequencing"]
related: ["concept-digital-transformation-1-0", "concept-single-instance-data"]
action: "Integrate operational data into common standards to create a single source of truth before building AI models."
outcome: "A reliable data foundation that prevents AI models from generating contradictory recommendations, ensuring long-term user adoption."
sources: ["tail1"]
sourceVaultSlug: "hbr-seg-tail1"
originDay: 1
articleStem: "hbr-tail-107-lenovo-ai-supply-chain"
sourceUrl: "https://hbr.org/2026/05/how-lenovo-built-an-ai-powered-supply-chain"
sourceTitle: "How Lenovo Built an AI-Powered Supply Chain"
---
# Fix data infrastructure before deploying AI

**Action:** Integrate operational data into common standards to create a single source of truth *before* building AI models.

**Do this because:** Resist the pressure to show quick AI wins. Spend the necessary time — even if it takes years — to integrate operational data from manufacturing, logistics, procurement, and fulfillment into common data standards and a unified architecture ([[concept-single-instance-data]]). This is the operational form of Phase 1 in [[framework-lenovo-two-phase-ai]] and what [[concept-digital-transformation-1-0]] actually required. It presupposes the data-engineering competence in [[prereq-data-standardization]] and the patience argued in [[contrarian-patience-over-speed]].

**Expected outcome:** A reliable data foundation that prevents AI models from generating contradictory recommendations ([[concept-broken-data-foundation]]), ensuring long-term user adoption ([[claim-ai-adoption-collapses-18-months]]). Pair with [[action-maintain-data-quality]] so the gains persist.