How agentic AI can connect operational data

Organisations are generating more operational data than ever but turning that information into timely decisions remains difficult when it is distributed across multiple systems.

But according to a new use case from Lufthansa Industry Solutions, agentic AI could provide a way to bridge these silos without requiring organisations to replace their existing technology infrastructure.

In How Agentic AI Enables Better Decisions in Complex Supply Chains, Lufthansa Industry Solutions proposes using an AI agent as an orchestration layer across ERP, logistics, IoT and business intelligence systems, allowing employees to interrogate information through natural language and receive contextualised analysis and recommended actions.

“Agentic AI becomes real when it moves from answering questions to getting work done,” said Dr. Stanislaw Schmal, Director Data & AI at Lufthansa Industry Solutions. “Whether coordinating maintenance activities, handling customer requests, or managing operational disruptions, AI agents are already helping organizations improve speed, consistency, and productivity.”

In a supply-chain environment, applications could include identifying material shortages earlier, assessing alternative transport routes, optimising inventory and simulating different scenarios. The broader principle is to connect information already held across an organisation, allowing relationships and emerging risks to be identified without the same level of manual data gathering and analysis.

However, moving towards AI agents that play a greater role in operational decision-making also raises questions around trust and accountability. Schmal emphasises there must be integration with existing business rules, governance mechanisms and role-based authorisation models.

“Trust is not built by the AI itself. It is built through clear governance, human oversight, and measurable results,” Schmal said. “My simple rule: if an AI agent can consistently improve a KPI such as turnaround time, response time, cost, or productivity while remaining transparent and auditable, organizations will trust it with more responsibility.

“The future of agentic AI is not pure autonomous decision-making. It is accountable decision-making at scale.”

Photo by Dennis Gecaj on Unsplash