AI Marking Software for Federations and Multi-Site Schools

If you lead a federation, a hard federation, or a multi-site secondary school, you already know that marking policy is one of the genuinely difficult things to standardise. Each site has its own history, its own department heads, its own inherited conventions, and often its own preferred mark schemes. Imposing a single rigid marking policy across the federation rarely works - staff resist it, the rollout takes a year, and the pilot site quietly carries on doing what it always did. AI marking software offers a different route: shared infrastructure that gives federation leaders the consistency they need without dismantling site-level autonomy.

This guide is for headteachers, executive heads, and central trust leaders evaluating AI marking tools for federation deployment. It covers the consistency problem honestly, what per-teacher pricing with unlimited marking actually buys you, and how onboarding works when the signatory sits at the centre rather than at a single site.

The Marking Consistency Problem Across Multiple Sites

Federations and multi-site schools live with a structural tension. Central leadership wants comparable outcomes - predictions, mock results, intervention triggers - that mean the same thing at every site. Site leadership wants to keep the autonomy that makes their school work in their local context. Marking sits exactly on this fault line. A grade 5 awarded at one site needs to mean the same as a grade 5 awarded at another, but the moderation structures that would guarantee this are expensive and politically delicate to set up.

Most federations end up with parallel moderation cycles per site and an annual cross-site standardisation meeting that produces good intentions and limited follow-through. The result is grade variation that the central team knows about but cannot easily fix. AI marking does not solve this on its own, but it shifts the baseline. If every site is using the same tool against the same uploaded mark scheme, you have a common reference point that did not exist before.

What One Price per Teacher Gives a Federation

The practical mechanism that makes deployment work is simple pricing. A GradeOrbit school account is one flat price per teacher, billed annually, and every seat includes unlimited marking and unlimited AI detection - no credits to buy, pool or ration, and no per-script charge. The commercial model matters because it removes a friction that kills most multi-site rollouts - the need for individual departments to find budget, raise POs, and approve spend every time staff want to use the tool.

For federation leaders, this also means certainty. The cost depends only on how many teachers have a seat, so it is known before the year starts, and a heavy mock season at one site costs nothing extra. You can then focus on which departments adopted quickly, which ones need more support, and where the workload-reduction story is actually landing, rather than on who is spending what.

Site-Level Autonomy Without Losing Central Oversight

The reason federations stall on technology rollouts is that the central spec usually overreaches. AI marking deployment works better when the central team owns the contract and the seat count, but each site keeps autonomy over which subjects pilot first, which mark schemes get uploaded, and how the tool sits inside their existing marking policy. The standardisation comes for free at the back end - every script across the federation is being assessed against the same exam-board criteria with the same model - without anyone at site level feeling that policy has been done to them.

This is also the right model for the academic conversation. Heads of subject across sites can compare AI-suggested grade distributions for the same paper, and that data becomes the input to a better moderation conversation than the federation has had before. The tool does not replace moderation. It gives moderation a starting point that everyone trusts.

Onboarding a Federation: DPA, Signatory, Retention Window

Onboarding GradeOrbit at federation level follows a structured signatory flow. The central signatory - typically the executive head, COO, or DPO - creates the federation account using a school email address. The Data Processing Agreement is reviewed and signed at this central level, covering all sites under the federation. URN is optional during signup. Once the central setup is complete, teachers are invited by email on the school's domain and can mark from day one, during a 30-day free trial with no card needed. The seat count and start date are chosen when the trial ends, with no charge until the signatory confirms.

The data position is designed for institutional comfort. A federation runs as a school account under that signed DPA: pupils' work is kept against their real names so teachers can come back to it, scans are removed when a marking run archives (90 days by default, adjustable from 7 to 365 days), and pupil records are wiped at the end of the federation's retention window (one to three years, three by default) or at offboarding. Only individual teacher accounts, outside a school agreement, discard uploaded work after marking. This means the federation DPO is signing off on a tool with a defined answer to what is held, for how long, and how it is deleted.

Talk to Us About GradeOrbit for Your School

GradeOrbit supports federation and multi-site deployment with central signatory onboarding, one price per teacher with unlimited marking, and the same data-minimisation architecture that protects student work at every site. If you lead a federation or a multi-site school and want to discuss how a deployment would work in your context, visit the GradeOrbit homepage and get in touch with our team. We are happy to walk through onboarding, DPA, and seat numbers in detail before any commitment.