Can ChatGPT make a school timetable?
3 min read
It's a fair question. AI chatbots like ChatGPT, Claude, Gemini and Copilot can write lesson plans, reports and letters, so why not a timetable?
For a small, simple timetable, a chatbot can produce a reasonable first draft. For a real school or college timetable, it can't be trusted to get it right. Here's why, and what to use instead.
What a chatbot can do
Give a chatbot a short list of classes, teachers and periods, and it will produce something that looks like a timetable. That's useful for:
- Sketching the structure of a school day
- Drafting a timetable for a club, a small department or a single class
- Explaining timetabling ideas or checking your reasoning
- Writing the letter that goes home with the new timetable
Why it can't build a real timetable
It doesn't check its work
A large language model generates text that is likely to fit the request. It doesn't systematically check every constraint. A timetable can look right and still have a teacher in two rooms at once, or a student in two lessons. In a timetable with hundreds of lessons, those mistakes are easy to miss and costly to find later.
The problem is too big to hold at once
A secondary school timetable involves hundreds of students, dozens of teachers, many rooms and thousands of possible slots. Every placement affects others. A chatbot works through the problem as text, and it can't reliably keep track of every interaction across the whole timetable.
It can't prove a timetable is impossible
Sometimes your rules can't all be met: there aren't enough qualified teachers at the times they're needed, or enough labs. A good timetabling tool tells you so. A chatbot will usually produce something anyway, quietly breaking a rule to do it.
It can't optimise
Among all the timetables that meet your rules, some are much better than others: fewer student gaps, fairer teacher days, staggered lunch. Finding the best one means searching and comparing a vast number of options. That's a different kind of computing from generating text.
What kind of AI can
The kind of AI that solves timetables is optimisation, also called constraint solving. Rather than generating a likely-looking answer, an optimisation solver:
- Takes every rule as an exact constraint
- Searches the possible timetables systematically
- Only returns a timetable that meets every hard constraint
- Scores each valid timetable against your goals and keeps the best one it finds
- Reports when no valid timetable exists
The same family of techniques plans airline crews, delivery routes and factory schedules, where a solution that looks right isn't good enough.
Where chatbots and solvers meet
The two kinds of AI work well together. A conversational interface can make a timetabling tool easier to use: describing a rule in plain English, asking why a lesson is where it is, or exploring a "what if". But the timetable itself should come from a solver that checks every rule.
How Timez works
Timez uses a constraint solver to build timetables. It takes your students, staff, rooms, subjects and availability, treats every hard rule as exact, and returns only timetables that meet them all. Then it works to cut student gaps, teacher gaps and lunch congestion, and reports how well it did. If your rules can't all be met, it tells you.
Related: hard and soft constraints in school timetabling and how to choose timetabling software.