How a giant's finance team cut 80% of the time on a repetitive task

Centalion (formerly Gunvor), a Geneva commodities trader that had revenue of US$ 144 billion in 2025, put an AI agent in its treasury, in partnership with Komgo.

Before, the team jumped between systems to check credit limits, look up data and retype each guarantee.

Having everything digitized wasn't enough: the bottleneck became the back and forth between screens.

Now the analyst asks, in plain language, whether there is a limit to issue or amend a guarantee. The agent fetches the data, applies the rules and leaves the operation ready for a person to approve.

According to the company, issuance time fell 55% and amendment time, 80%. Less repeated typing, less risk of error.

Almost every company has a process like this in finance: check a balance in one system, copy data from another and post it in a third.

List those steps, let AI prepare the operation and keep the approval with whoever is accountable for it. Centalion started with a single credit line before expanding.

The effort lies in organizing rules and access first, because the agent only follows what is written.

The numbers are the company's own, with no independent verification, and they apply to that first line. When the agent covers the whole area, the projection is to free up more than 60 days of work a year.

How many screens does your finance team open for a single entry?

Sources

Centalion, 8 Oct 2026. https://centalion.com/news/centalion-and-komgo-bring-agentic-ai-into-live-treasury-operations/

PYMNTS, 19 Oct 2026. https://www.pymnts.com/news/artificial-intelligence/2026/ai-agent-cuts-centalions-trade-finance-amendment-time-80/

Komgo. https://www.komgo.io/newsroom/centalion-and-komgo-bring-agentic-ai-into-live-treasury-operations