Before the solution was introduced, dispatch decisions across the pharmacy hub's network were handled manually, requiring the logistics team to evaluate hundreds of medication orders and determine the appropriate pharmacy location and shipping method for each one.
During the initial 4-6 week pilot, Claude enabled 85-90% of orders to receive a high-confidence automated dispatch decision, significantly reducing the number of orders requiring manual decision-making.
The remaining 10-15% of orders were handled through a human-in-the-loop workflow, with the AI-generated decision presented as a recommendation for a logistics specialist to review and approve.
The pilot also provided valuable operational feedback. By analyzing real-world cases that required human intervention, the team identified additional factors affecting dispatch decisions and defined the next round of improvements to increase decision precision and expand the share of orders that can be automated.
