Coupang and Nvidia build AI factory to support logistics and delivery operations

Coupang built an AI factory with Nvidia, pairing its Intelligent Cloud with Nvidia's DGX SuperPOD so operations teams can develop and test models without waiting for central approval. The setup supports Rocket Delivery and route optimization work.

Categorized in: AI News Operations
Published on: Mar 31, 2026
Coupang and Nvidia build AI factory to support logistics and delivery operations

Coupang and Nvidia Build AI Factory for Logistics Operations

Coupang announced on March 17 that it has built an "AI factory" with Nvidia to accelerate model development across its e-commerce and delivery operations. The system pairs Coupang's Intelligent Cloud with Nvidia's DGX SuperPOD, creating a self-service environment where teams can build and test models across global operations.

The partnership addresses a core operational challenge: speed. Operations teams can now develop AI models without waiting for centralized approval or infrastructure provisioning. Coupang said the system supports its "Rocket Delivery" service and positions the company as a launch partner for Nvidia's Dynamo software.

What This Means for Operations Teams

For operations professionals, this setup removes friction from AI adoption. Teams can experiment with models for route optimization, demand forecasting, and warehouse automation without lengthy procurement cycles. The self-service model lets operations staff test ideas quickly and measure results before broader rollout.

This approach differs from traditional enterprise AI deployments, where centralized data science teams create models for operations to implement. Here, operations teams have direct access to tools and infrastructure.

Financial Context

Morgan Stanley lowered its price target on Coupang to $29 from $31 in March, citing improving operating metrics in Korea and easing regulatory concerns tied to a prior data breach. Barclays raised its target to $24 from $23 after fourth-quarter results.

Coupang reported Q4 earnings per share of negative 1 cent against a 4-cent consensus estimate, with revenue of $8.835 billion versus $9.09 billion expected.

Building Operational Capability

For operations managers considering similar infrastructure investments, the Coupang-Nvidia model illustrates a practical approach: pair cloud services with specialized compute hardware, then grant teams direct access rather than creating bottlenecks through centralized governance.

Learn more about AI for Operations and explore how operations professionals can lead AI implementation in your organization with our AI Learning Path for Operations Managers.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)