Atmora Tech

Industry

Student systems that hold up at enrolment

Student and learning systems that survive enrolment week and the statutory return that follows it.

Context

Technology in Education

We build learning platforms, student administration integrations, admissions and enrolment journeys, and the reporting that funding depends on. The calendar drives everything: an institution has a handful of days each year when every student logs in at once, and a small number of fixed dates when returns must be filed correctly or funding is affected.

Most of the difficult work is integration with a student information system that will not be replaced, plus safeguarding and children's data obligations that constrain design. Artificial intelligence features are worth discussing only after the basics hold: an adaptive tutor is not the fix for a timetable that puts two cohorts in the same laboratory, and we would rather spend your budget on the timetable.

48,000
concurrent users handled during a results-week enrolment window
9 hours to 20 min
roster synchronisation lag after moving to OneRoster
0
rejected statutory returns across four consecutive census dates

Challenges

What actually keeps this sector awake

Not generic disruption — the specific constraints that shape every technology decision here.

  • Enrolment peak in a four-day window

    Annual load concentrates into results week, when the same infrastructure that idles all summer must take every returning student, every applicant and every parent at once. Failure here is visible to the whole institution within minutes.

  • Student information systems that resist integration

    The SIS holds the authoritative record behind batch interfaces and a licence that limits access. Learning tools each build their own partial copy of the roster, and by term three no two systems agree on who is enrolled.

  • Safeguarding and children's data by design

    Age-appropriate design expectations, parental consent, data minimisation and safeguarding escalation paths have to be in the product rather than in a policy document. Analytics and third-party tooling need review before they ship, not after.

  • Statutory returns with unforgiving validation

    Census and funding returns are validated against rules that reject the whole file for a handful of records. Institutions discover data quality problems days before a deadline because nothing checks continuously against the same rules.

Approach

How we address them

  • Standards-based learning integration

    LTI 1.3 tool integration with proper deep linking and grade passback, and roster synchronisation through OneRoster so enrolment changes propagate rather than being re-keyed. One authoritative roster, many consumers.

  • Capacity engineering for the enrolment window

    Load modelled from last year's actual traffic shape, tested against production-sized data, with queueing and graceful degradation designed in so a slow downstream system delays a page rather than taking the service down.

  • Safeguarding workflow with a real audit trail

    Concern recording, escalation and access control designed with designated safeguarding leads, with strict role separation, tamper-evident history and retention rules that match the institution's obligations.

  • Continuous validation of statutory returns

    Return rules implemented as tests that run daily against live data, so failing records surface in October rather than the night before submission, with an owner assigned per exception category.

Start a project

Tell us what is
breaking.

We reply within one working day, and the first call is with an engineer who would actually work on it — not an account manager. If we are not the right studio for the problem, we will say so on that call.

Start a project