INTELLIGENT AUTOMATION OF A VOCATIONAL COLLEGE TIMETABLE: A HYBRID MODEL OF OR-TOOLS/ILP AND AN LLM ASSISTANT

Abstract

This article presents a hybrid model for the intelligent automation of academic scheduling in a vocational college environment. The relevance of the study is explained by the fact that traditional manual timetable creation requires considerable time and organizational effort, often leads to conflicts, and makes it difficult to simultaneously balance teacher workload, classroom resources, group study time, shift-specific constraints, and additional administrative requirements. The purpose of the study is to improve timetable generation quality by integrating an OR-Tools-based constraint satisfaction model with an LLM assistant capable of working with natural language, while also increasing system flexibility and responsiveness to changes. The research employs system analysis, constrained optimization, scenario modeling, comparative evaluation, and applied design methods. The results demonstrate that the hybrid approach significantly reduces timetable generation time, more accurately reflects teachers’ time preferences, decreases the number of student idle gaps, supports more efficient allocation of classroom resources, and simplifies rapid schedule adjustments. The scientific novelty lies in combining an LLM layer that transforms natural language requests into formal constraints with the OR-Tools/ILP computational core. The practical significance of the proposed solution is that it enables college administrators to improve planning quality, use resources more efficiently, and ensure a more stable and adaptable organization of the educational process.