What AI Changes, and Doesn’t, in Exception Handling

AI-native platforms create an opportunity for agents to act first on supply chain exceptions and document the response as it unfolds, rather than after the fact, according to Henry Ames, vice president of corporate strategy at Zoomlogi.

Speaking with Pharmaceutical Commerce at LogiPharma USA 2026, where he joined the panel "From SOP to Audit-Ready: AI-Enabled Exception Management Across the Pharma Supply Chain," Ames says exception handling today relies heavily on experienced staff who know how standard operating procedures (SOPs) vary by product, customer and trade lane.

"It requires a very manual process and is hugely error-prone," he says.

Documentation often comes well after the fact, as staff log into multiple portals and call couriers and shipping sites to piece together what happened. With AI-native platforms, Ames says, "there's a real opportunity to first and foremost digitize those workflows, allow the agent to be the first point of action to go out to collect the information, and then actually have that documentation get done as the situation is unfolding."

The underlying workflow follows the SOP whether a shipment is clinical or commercial, Ames notes, even though the networks differ. Clinical shipments tend to be lower-volume, higher-value parcels, sometimes one of one for autologous therapies. Commercial products move from manufacturers through wholesalers and distributors to pharmacies, hospitals and clinics. What changes is likely the level of automation, which he sees as potentially higher on the commercial side. Agents can do much of the background work, he says, and present the information to a person for the judgment call.

Watch the first installment of Ames’ interview:

1. Why AI’s Clearest Supply Chain Wins Are in the Basics