Less preparation. More informed review.

Use AI where it is genuinely helpful: gathering context, comparing records, preparing a first pass, and explaining what deserves attention.

Walk through this workflow
What changes

Every suggestion lives inside the same permissions, rules, review steps, and audit history as the underlying workflow.

Schoolbooks / AI-assisted workReview workspace

Why did fee income move this week?

Prepared answer

The movement is concentrated in two receipt batches and one approved fee adjustment.

3 linked records No entry changed

The workflow

Follow the work, not the org chart.

Each stage carries its evidence and status forward, so the next person receives context instead of another handoff problem.

  1. 01

    Ask in context

    Begin from a payment, reconciliation, account, period, or approved report—not a blank chat box.

  2. 02

    Gather permitted evidence

    Retrieve only the records and documents the current user is allowed to access.

  3. 03

    Prepare a transparent answer

    Return the proposal, linked evidence, reasoning, and uncertainty together.

  4. 04

    Hand control back

    Let a person approve, correct, dismiss, or request more evidence before any material action.

  • Finance staff repeatedly search for the same evidence across systems.
  • Reviewers receive totals without a useful explanation of what changed.
  • Important exceptions are hidden inside long operational lists.
  • Generic chat tools lack the permissions and accounting context of the school.

Permission-aware context

The assistant cannot reveal records a user could not open directly.

Evidence-linked answers

Financial explanations link back to the entries used to create them.

Human approval

AI output remains a proposal until an authorized workflow advances it.

The boundary

Useful automation should make accountability clearer.

AI may

  • Search, compare, summarize, and prepare.
  • Explain a suggestion in school-specific language.
  • Route uncertainty to the right review queue.

AI must not

  • Act as an unaccountable autonomous bookkeeper.
  • Present unsupported answers as financial fact.
  • Operate outside configured access and approval rules.
Continue exploring

Students, guardians & admissions

One learner record from enquiry to enrollment.

Explore Students, guardians & admissions