Educator-in-the-Loop: AI-Assisted LaTeX Workflows for Rapid Development of Technical Lecture Materials in Engineering Education

This paper presents an educator-in-the-loop framework for integrating generative AI with LATEX authoring in civil engineering education, specifically focusing on fluid mechanics and hydrological engineering. The proposed workflow enables instructors to use AI tools such as ChatGPT and Gemini by providing input in the form of handwritten notes, scanned PDFs, lecture slides, or images. These tools generate initial LATEX drafts, which are then refined and validated by the educator using editors like Overleaf. This collaboration streamlines the creation of high-quality and technically accurate teaching materials, allowing educators to focus on pedagogy rather than formatting. Four representative case studies, deriving the Bernoulli equation, unit hydrograph plotting, application of Manning’s formula and schematic representation of pipe flow, demonstrate practical implementation of the workflow. Each case highlights how AI accelerates the writing while maintaining instructional integrity through expert oversight. The educator-in-the-loop model enhances productivity, ensures consistency across materials, and supports reproducibility, making it a promising strategy for modernizing content development in STEM education.