Next year's A–F grade for every campus, on its own model's cuts and with the range it could really land in. Then the checks that stand between an Infinite Campus export and an ADE submission. Built to your network's files, and run on your own server.
Seven invented campuses. The arithmetic is the real model's, fitted on every graded Arizona school. Hover a campus for its numbers.
Arizona grades a K-8 campus, a 9-12 campus and a hybrid on different weights, and the A line is not in the same place. A K-8 campus needs 84 for an A; a 9-12 or hybrid campus needs 82. Put every campus on one scale and it misgrades about one K-8 campus in four.
The K-8 and 9-12 sides are scored on their own models, then averaged by enrollment at the start of spring testing. The result is graded on the 9-12 cuts.
What it covers: K-8, 9-12 and hybrid campuses. Alternative schools are graded on a different model and are quoted separately.
Four jobs a data office does every year by export and spreadsheet. Each is rebuilt as rules with a test for every rule, so a result can be traced back to the row and the rule that made it.
Reads the state results, ACT Aspire, ACT and AP files as the vendors send them, matches every row to a student in IC by state ID and name, and checks the score, the level, the grade and the test date. Clean rows become an import-ready file; the rest come back with a reason.
Pulls each senior's evidence from IC: dual enrollment and core-course grades, CTE, ACT, AP, ASVAB, industry credentials, work-based learning. Scores it on ADE's current point table and rolls it up to the campus score, one row per senior, ready for the ADEConnect report.
If your network raises a semester grade for a proficient test result, every rule in your written plan becomes a rule with a test: which tests, which courses, the caps, opt-outs, incomplete courses. Every changed grade carries the rule that changed it.
Eligibility, observation weighting, the share of each pool and proration, straight from your adopted plan. The plan's own worked example is one of the tests, and the pool sizes are settings, because they depend on what is funded.
| # | State ID | Last | First | Grade | Content | Scale | Level | Test date | Result |
|---|
| State ID | Last | First | Grade | Already in IC |
|---|
Click any cell and change it: the checks run again as you type. The scale (400 to 800, Proficient from 600) is invented for the demo. In a real build, a file of planted errors ships with the checks and every one must be caught, with no clean row flagged; a check that cannot fail is not a check.
A web dashboard on your own server, behind your own sign-in. Every table behind it also comes out as a plain CSV, so the tools you already use can read it.
Exports you already make, as CSV or Excel, in whatever layout your IC produces. The build is fitted to your layout, not the other way round.
Made for your network, one piece or all of them. Student records never leave your building.
A header row and a blank copy of each export. The build runs against invented students shaped exactly like your files, like the demo on this page.
The code, the tests and the dashboard install on your server and read your own exports. Each cycle is one command, written down in the run book.
One fixed price, agreed before any work starts. No subscription, nothing that renews on you. A new rules year is a quoted update, never an automatic charge.
For one network: the forecast, all four checks and the dashboard, built to your files. The exact figure is agreed before any work starts, and a single piece is quoted on its own. [email protected]
An Arizona classroom teacher with an M.S. in Computer Information & Decision Management, and a data analyst before that. Every rule here is written from the state's published business rules or your own plan, and every rule has a test that fails when the rule is broken.