Tier 1 · FourKites Concepts

Agent Operating Procedures (AOPs)

An Agent Operating Procedure (AOP) is an executable playbook that defines exactly how an AI agent handles a specific supply chain situation: which data to gather, which intelligence to consult, which action to take, which channels to use, and when to escalate.

Why It Matters

Decision trees are brittle. Prompt engineering is fragile. AOPs are neither. They combine the flexibility of natural language with the precision of coded logic, powered by the FourKites Graph. Every step is defined, every decision point is explicit, every outcome is traced.

The FourKites Perspective

FourKites AOPs are unique because they connect every step to the Graph and Digital Twins. A generic workflow builder says 'check if the shipment is late.' An AOP on the Graph says 'check the Shipment Twin, pull the carrier's historical reliability on this lane, evaluate root cause using network intelligence, and select the resolution path with the highest success rate across 2,499 prior outcomes.'

Frequently Asked Questions

What is the difference between an AOP and a traditional workflow?
Traditional workflows follow static if-then rules. AOPs draw on live intelligence from the FourKites Graph at every decision point. The same AOP produces different actions depending on the carrier, the lane, the season, and what has worked before.
Who builds AOPs?
Sophie, FourKites' AI Developer Agent, builds AOPs from natural-language descriptions. Describe your supply chain process in plain English. Sophie translates it into a production-ready AOP connected to the Graph and Digital Twins.
Are there pre-built AOPs?
Yes. Loft ships with a library of pre-built AOPs covering common supply chain exceptions (late load follow-up, supplier document processing, appointment scheduling, WISMO resolution, POD collection). Each has been validated across Fortune 500 deployments.

Related PLATFORM PAGES

See how this concept powers autonomous operations.
Talk to an outcome advisor