Intelligent Agents Based on Large Language Models in Transport Management: A Comparative Analysis of Approaches to Development and Implementation

The paper addresses the “information gap” between objective telemetry data and primary documentation in logistics and road construction, which arises from the growing volume of unstructured information and a shortage of qualified personnel. Two projects aimed at creating intelligent agents to bridge this gap in transport management are examined. Based on a comparative analysis of two implemented solutions, the study identifies common architectural principles. The first project (technical proposal stage) develops an AI‑powered dispatcher assistant that automatically collects and matches requests from external channels with an internal TMS system via deep API integration. The second project (working prototype) implements a workflow on the low‑code n8n platform that extracts, normalises, and compares GLONASS PDF data with accounting records. Both agents share a unified five‑stage cognitive agent model and a hybrid architecture combining a large language model with deterministic logic. The two solutions differ fundamentally in their integration paradigm: deep API integration versus low‑code assembly. This difference determines development speed, total cost of ownership, and scalability. The low‑code prototype achieves a payback period of less than one month, while deep integration offers greater customisation and ERP connectivity at significantly higher initial cost. Criteria for selecting an architectural approach are formulated based on the maturity of a company’s IT landscape. These criteria enable a strategic choice between deep integration and low‑code development depending on business priorities, infrastructure readiness, and available resources.