Switching from a legacy scheduler

Keep your SIS. Replace the scheduler.

Your student information system is the record of your school — enrollment, grades, transcripts, state reporting. Nothing here suggests replacing it. The scheduling module bolted onto it is a different question.

This is not an SIS migration

SIS migrations are among the most disruptive projects a school can undertake. This is not one of them, and if anyone pitches replacing your student information system as a way to fix scheduling, be suspicious.

Why the built-in module frustrates people

  • A result without a reason. They rarely fail outright; they succeed opaquely. You get 340 unfilled requests and no account of why any one of them failed, so the work turns forensic.
  • The vocabulary is the product's, not yours. Constraints are codes, flags and weights documented nowhere. Schools cope with folklore — "set the balance weight to 3 or it ignores room caps."
  • Administrators described a “welcome departure from ‘clunky’ and ‘archaic’ scheduling systems, which slowed and complicated the process.” Columbia CPRL, About Time: Master Scheduling and Equity, p.17 (2021)
  • You get one attempt. Most run against live data and overwrite the previous result, so "what if we optimized for balanced sections instead?" costs you the schedule you already have.
  • Nobody is improving it. Scheduling is one module among dozens competing with compliance reporting, gradebook and attendance. It is nobody's flagship, and the interface shows it.

Columbia's landscape analysis of eight scheduling tools found seven of eight alert on course conflicts — but only four of eight when a student is scheduled into a course they already took, or is missing one they need to graduate.

Connect the student information system your school already runs on.
Connect the student information system your school already runs on.

The round trip

  1. Connect

    An account administrator connects your SIS. Credentials are encrypted at rest and the connection is scoped to what scheduling actually needs.

    Admin-only, by design

  2. Pull

    nOS imports courses, sections, teachers, rooms, students and course requests. A pull report tells you exactly what came across and what it could not read, rather than silently dropping rows.

  3. Clean

    The Data Cleaner surfaces the problems that exist in every real dataset — duplicate course codes, requests pointing at retired courses, teachers loaded past the number of periods in a day. You fix them before anything is built.

  4. Build and refine

    Generate across five optimization goals, compare scenarios side by side, drag sections on the Board with live conflict detection, and type constraints in plain English. Every edit is staged and previewed before it commits.

  5. Preview the write-back

    Before anything is sent, an administrator sees exactly what nOS would write into the SIS. This is the step that makes the round trip safe — you approve a diff, not a black box.

    Nothing is written until you approve it

  6. Push

    The finished schedule goes back into your SIS. Every push is recorded with counts, outcomes, and the exact schedule snapshot it came from.

Connector coverage

nOS speaks to 19 student information systems, and 18 of the 19 can receive a schedule back — Clever is pull-only. Eleven of them ride a shared, live-tested OneRoster 1.1 connector.

The rest are vendor-specific connectors, and nOS shows each one's status inside the app rather than presenting them all as equivalent — so you always know exactly what you are working with. Tell us which system you run and we will confirm where yours stands before you plan around it.

See the full list of systems →

The setup guide: 600 students, 126 sections, 3,284 requests, preflight passed.
The setup guide: 600 students, 126 sections, 3,284 requests, preflight passed.

Side by side

  A built-in SIS scheduler nOS
System of record Your SIS. Correctly so. Still your SIS. nOS builds; it does not replace.
Unfilled requests Usually a count, sometimes a code. Each one carries the specific reason it could not be placed.
Trying an alternative Typically overwrites the current result, so it rarely gets tried. Scenarios saved and compared side by side; promote the winner.
Expressing a rule Codes, flags and weights whose interactions are largely undocumented. Type it in plain English; nOS compiles it into a solver constraint.
Manual adjustment Form-based edits, often without live validation. Drag on a resource × period Board; validity shown before you let go.
Writing back Native — it is the same database. Preview the exact diff, then push. Supported on 18 of 19 connectors.
Asking a question of the schedule Run a report, read the report. Ask in plain English and get an answer from the live schedule.
Cost Bundled — which makes it feel free and makes it hard to evaluate. $8 per enrolled student per year, stated publicly. Pricing.

"A built-in SIS scheduler" describes the common pattern across the category rather than any single named vendor. Check the rows against your own module.

Adopting this safely

  • Run it in parallel first. Pull your data, build a schedule in nOS, and compare it against the one your current module produced. Nothing is written back unless you push it. This costs you a few hours and tells you more than any demo.
  • The write-back is previewed. An administrator sees the exact set of changes nOS would make before any of it reaches your SIS.
  • Your SIS remains authoritative. If a build goes sideways, you have not lost anything — the record still lives where it always did.
  • Exports work regardless. Student schedules, rosters, teacher and room schedules and unfulfilled-request reports all export to CSV, Excel and PDF, so you are never dependent on the connector to get your work out.

Run it in parallel. Judge it on the result.