Schools deserve better systems than the ones they have.

The software schools are sold is usually a set of modules that were built separately and joined afterwards. Each holds its own copy of the student list, and keeping them in agreement becomes a job somebody does every week.

We thought the sensible thing was to start from the record rather than the modules: one student, one enrolment, one set of guardians, and every part of the product reading from it.

It should remove work, not move it. If a system needs someone to keep it updated, it has added a job rather than replaced one.

A school is not an office. Registers are marked while thirty children wait. Fees are collected at a counter with a queue. Screens have to work at that speed.

Separation is not a setting. One school must never be able to reach another school’s data, and a parent must never reach another family’s. That belongs in the foundations, not in a checkbox.

Say what it does not do. Our own feature page lists what is still being built. A school that finds out during implementation was told too late.

The product is written as one application over one database, with the tenant boundary enforced where data is read rather than remembered by each screen. Sensitive actions are written to an audit trail. Tests cover the boundaries that would be expensive to get wrong.

We would rather ship a smaller product that is correct than a longer feature list that is approximately true.

Private schools, international schools, preschools and multi-campus groups — the institutions large enough that spreadsheets have started to hurt, and small enough that enterprise software would be absurd.