Teachers do not need more software. They need their evenings back. - Lesson planning is the standout: blank page to a week's structured plans in an afternoon. - Grading helpers work, with a human in the loop on anything that lands on a student's record. - Check where student work goes when a tool processes it. Free tools are often free because the data is the product. Have that conversation with IT first. - Start small: one prep-heavy subject, one tool. If it saves three hours a week, expand. - Differentiation is the quiet killer feature: one lesson rewritten at three reading levels saves the most painful prep. I compare AI tools across 40+ professions including education, and PrivateLLM deploy sets up a private LLM on AWS for $50 plus usage for districts that want student data to stay in-house.