Amazon Connect Decisions: Agentic AI Aims to Turn Supply-Chain Visibility into Action

Amazon Web Services introduced Amazon Connect Decisions, an agentic supply-chain planning and intelligence tool that consolidates more than 25 specialized supply-chain functions into AI “teammates.” The system links to existing ERP and other data sources, offers both visual dashboards and a chat interface, and claims to shrink planning cycles from days to hours. Launched in 2024, the platform builds on AWS Supply Chain (2022) and the SCOT foundation model. Early adopters include Wells Vehicle Electronics and TVS Motors, with over ten beta customers testing the service since late 2023.

Amazon Launches AI Hiring and Supply Chain Tools to Cut Interview Time

Industry commentary notes that a decade of investment in real-time tracking, control towers, and event platforms has delivered unprecedented visibility, but visibility alone does not prescribe actions. Effective AI must embed decision logic—rules, priorities, and authority—that converts signals into responses. Without integrated execution pathways, many AI pilots remain advisory, surfacing problems without delivering concrete resolutions.

Visibility Isn’t Decision-Making in Supply Chain AI

AWS leverages its experience managing more than 400 million SKUs to train Connect Decisions. The platform can ingest thousands of alerts, prioritize tasks, and surface “invisible patterns” that help companies reduce unnecessary inventory. In beta trials, the AI teammates have traced root causes, suggested remediation steps, and learned from each planner’s action, spreading those learnings across the organization.

SupplyChainDive

• Ozgur Dogan, VP of AWS Supply Chain, emphasized that the system continuously learns from planner actions, spreads improvements organization-wide, and maintains full transparency so users stay in control.

SupplyChainDive

• Abhishek Gupta, Director of Forecasting and Labs for Amazon SCOT, described SCOT as a model that decides what to buy, where to place inventory, and how to ship, using historical data—including holidays and promotions—to sharpen forecasts. Both executives stress that Connect Decisions adapts to existing processes rather than forcing operational change.

SupplyChainDive

Analysts argue that AI recommendations lack impact unless tied to clear authority, workflow, and context. They warn that “more alerts do not equal better decisions” and that the true value lies in a control system that links visibility to decision logic, workflow, and feedback. Without such integration, AI tools risk remaining “smarter alerts” rather than autonomous decision engines.

Visibility Isn’t Decision-Making in Supply Chain AI

AWS positions Connect Decisions as an autonomous decision-maker capable of institutionalizing choices across an enterprise. In contrast, the broader industry view highlights that many AI solutions remain advisory, lacking embedded execution authority. This tension reveals a gap between the platform’s promised autonomy and the practical need for defined decision pathways.

Visibility Isn’t Decision-Making in Supply Chain AI

• “Every variance, invisible patterns — the system learns and gets better,” — Ozgur Dogan, VP, AWS Supply Chain

SupplyChainDive

• “And a supply planner can make an action, take an action, the system learns from it and spreads [it] to everyone.” — Ozgur Dogan, VP, AWS Supply Chain

SupplyChainDive

• “So, SCOT is deciding what items to buy, where inventory should be placed, and how it is stored throughout our global fulfillment network, and after, how do we ship it to our customers,” — Abhishek Gupta, Director of Forecasting and Labs, Amazon SCOT

SupplyChainDive

• “The value is in knowing which exceptions deserve action.” — Logistics Viewpoints analysis

Visibility Isn’t Decision-Making in Supply Chain AI

• “A control system connects visibility to context, decision logic, workflow, and feedback.” — Logistics Viewpoints conclusion

Visibility Isn’t Decision-Making in Supply Chain AI

AWS says that once teams adopt Connect Decisions, the platform can institutionalize decisions and scale them organization-wide without altering existing workflows. Future road-maps include deeper ERP integration, expanded AI teammate capabilities, and broader rollout of beta programs, aiming to move from visibility-only dashboards to fully orchestrated, decision-driven supply-chain operations.

SupplyChainDive