About nOS — nOS

Master scheduling is one of the highest-stakes jobs in a school, and one of the least supported. It decides which students get which courses, which teachers carry which loads, and which requests quietly go unmet — and in a lot of schools it is still assembled by hand, by one person, over weeks.

That job has an unusual shape. It is genuinely hard combinatorial math, so it wants a real solver. But it is also full of judgment that no solver can make — which singleton to protect, which teacher is coming back from leave, which two students should not share a room. The tools schools are handed usually pick one side. A spreadsheet gives you all the judgment and none of the math. A legacy scheduling module gives you a black box that produces a result you cannot interrogate.

nOS is built on the premise that you should get both: a solver that places thousands of course requests in minutes, and a workspace where you can see every decision it made, disagree with it, and change your mind — without starting over.

What we believe

The schedule should be legible

Every unplaced request in nOS has a reason attached. Not a status code — an actual explanation of what stood in the way. If software is going to tell a student they cannot take Physics, it owes the registrar an answer for why.

Nothing should apply itself

Edits in nOS are staged, not committed. You see the before and after and the fulfillment numbers a change would produce, then you decide. That applies to the AI too — it proposes, shows its work, and waits. The person accountable for the schedule stays the person who approves it.

Small schools are not small versions of districts

A 300-student school has genuinely harder scheduling math than a 3,000-student one — more singletons, fewer sections to move things into, less slack everywhere. What it does not have is a scheduling department. Most of this category is priced and shaped for districts. We build for the other end.

Student data is not ours

nOS operates as a school official under FERPA. Student data is never sold and never used to train AI models. It is hosted in the United States and encrypted in transit and at rest. The security page spells out exactly how, down to our subprocessors.

The company

nOS LLC is a small software company registered in North Carolina. We build three things: nOS, the scheduling platform; nOS AI, the co-pilot built into it; and the nOS AI Assistant, a separate standalone app that reads a schedule and answers questions about it without changing anything.

We work directly with the schools that use nOS. When a registrar tells us the refine step is confusing, that conversation tends to show up in the product within the week. That is the main advantage of being this size, and we would rather keep it than lose it.

Follow along