Errors in pharmacy medication directions, such as incorrect instructions for dosage or frequency, can increase patient safety risk substantially by raising the chances of adverse drug events. This study explores how integrating domain knowledge with large language models (LLMs)—capable of sophisticated text interpretation and generation—can reduce these errors. We introduce MEDIC (medication direction copilot), a system that emulates the reasoning of pharmacists by prioritizing precise communication of core clinical components of a prescription, such as dosage and frequency. It fine-tunes a first-generation LLM using 1,000 expert-annotated and augmented directions from Amazon Pharmacy to extract the core components and assembles them into complete directions using pharmacy logic and safety guardrails. We compared MEDIC against two LLM-based benchmarks: one leveraging 1.5 million medication directions and the other using state-of-the-art LLMs. On 1,200 expert-reviewed prescriptions, the two benchmarks respectively recorded 1.51 (confidence interval (CI) 1.03, 2.31) and 4.38 (CI 3.13, 6.64) times more near-miss events—errors caught and corrected before reaching the patient—than MEDIC. Additionally, we tested MEDIC by deploying within the production system of an online pharmacy, and during this experimental period, it reduced near-miss events by 33% (CI 26%, 40%). This study shows that LLMs, with domain expertise and safeguards, improve the accuracy and efficiency of pharmacy operations.

A robotic mobile fulfillment system (RMFS) performs the order fulfillment process by bringing inventory to workers at pick-pack-and-ship warehouses. In the RMFS, robots lift and carry shelving units, called inventory pods, from storage locations to picking stations where workers pick items off the pods and put them into shipping cartons. The robots then return the pods to the storage area and transport other pods. In this article, we consider an item assignment problem in the RMFS in order to maximize the sum of similarity values of items in each pod. We especially focus on a reoptimization heuristic to address the situation where the similarity values are altered so that a good assignment solution can be obtained quickly with the changed similarity values. A constructive heuristic algorithm for the item assignment problem is developed, and then, a reoptimization heuristic is proposed based on the constructive heuristic algorithm. Then, computational results for several instances of the problem with 10-500 items are presented. We further analyze the case for which an item type can be placed into two pods.