Education Administration /Customer-service agent live field guide · 8 min

Enrolment Nudge Drafter

Spots admitted students stuck between offer and start and drafts warm, personalised reminders for each outstanding step — for staff to approve and send, targeting the "summer melt" that quietly loses tuition.

Open the live demo

theater/demos/edu-admin_enrolment-nudge-drafter.html · sandbox · read-only

Open

FIG. 1

The live demo, running on fabricated data. Open it to step through the full flow — every output is shown for a person to approve before anything happens.

How it would work

Reads the stalled enrolment message and the student record, works out which steps are still outstanding, drafts a warm nudge that names them — and flags anything sensitive for a staff member to approve before it sends.

Input 01

The message + the student record

A help message in the enrolments inbox from an admitted student, their partial enrolment form and attachments (USI confirmation, prior results), and what the student record still shows as outstanding.

Agent 02

Classifies, extracts, drafts

Classifies intent and urgency against the academic calendar, routes it, pulls the course, intake and outstanding steps from the record, and drafts a short reminder that names each step and where to get help.

Output 03

A draft, with the sensitive bit flagged

A ready nudge with its working shown — and any sensitive mention, like fee hardship, flagged for escalation rather than auto-sent. A staff member approves, edits or reassigns before the student ever hears from it.

Where it works well

It catches the students who stall silently, and writes the chase you never have time to write.

Best for an admissions officer or student success coordinator working a backlog of part-finished enrolments in the offer-to-census window.
At a few hundred admits an intake, it surfaces the stalled ones and drafts each nudge in seconds — chasing every incomplete step by hand never scales.
The recaptured hours go into the students with real barriers — affordability, conflicting work, language — not into typing the same "you're nearly there" email forty times.

The students you lose to "summer melt" rarely complain — they just go quiet mid-enrolment, with a photo ID upload or an LLN pre-training review left undone, and the gap between offer and census closes around them while no one notices.

Where it works badly

It writes a confident, friendly nudge even when the record it read is wrong.

Weak when the message hints at something the record can't see — a disability, a hardship, a complaint. It must flag and escalate, not smooth it into a generic nudge.
If your enrolment status is updated in batches days late, the tool nudges against a stale picture and erodes trust with the exact students you're trying to keep.

The honest test

If you can't say, right now, how fresh the "outstanding steps" field is for a given student, this tool sends a faster wrong reminder, not a safer one.

If the student record says "photo ID outstanding" but the upload actually landed yesterday and wasn't reconciled, the tool cheerfully chases a student who already did the thing. The draft reads warm and correct; the data behind it isn't. That is the trap.

What it doesn't do — and shouldn't

It drafts the nudge. A person approves it. Sensitive cases it escalates, never sends.

WHAT IT DOES

Surfaces the outstanding steps it read and where each came from

Drafts the reminder naming each step and the help path

Flags sensitive mentions — fee hardship, distress, withdrawal — for staff

WHAT IT WON’T

Send any message to a student on its own

Decide a fee waiver, payment plan or subsidy eligibility

Action a withdrawal, deferral or hardship case

An enrolment message often carries personal and sometimes sensitive information — protected under the Privacy Act and the Australian Privacy Principles — and a fee or hardship reply can shape whether a student stays. A wrong or tone-deaf auto-send is a welfare and reputational risk, so the accountable staff member stays on every message that goes out.

What your data has to look like

A live enrolment status per student, and the message itself — not last week's export.

Typical readiness

across orgs we see, before the first job

Current enrolment status per admit

Needs shaping

The enrolment inbox and its attachments

Usual weak point

A routing map for the inbox

Needs shaping

Verified USI and identity on file

Usually ready

The academic calendar with census dates

Usually ready

The real first job

The weak point is almost always the enrolment status itself — split across a student management system, a forms tool and an inbox, reconciled by hand and lagging reality by days. Getting that status live and trustworthy is usually the real first job, larger and more valuable than the drafting layer on top. Once the record is current, every nudge after that is right by default.

Right fit if…

You work a real offer-to-census funnel — admits stall between offer and start

A few hundred-plus admits an intake, mostly the same handful of outstanding steps

Enrolment status is current enough to trust on a given day

Online or remote cohorts where you can't catch students in a corridor

Walk away if…

Your enrolment status is a weekly export reconciled by hand

Most stalled enrolments are one-off, complex hardship or welfare cases

You want it to send chases unsupervised, with no one reading them

You need it to decide fee waivers, subsidies or withdrawals for you

Open questions

The worried-buyer questions, answered straight

It can draft a wrong nudge — which is exactly why nothing sends on its say-so. It shows the outstanding steps it read and where each came from, drafts the reminder, and a staff member approves, edits or reassigns before anything reaches the student. Anything sensitive — fee hardship, distress, a withdrawal hint — it flags for escalation rather than drafting a breezy chase. The person stands behind every message that goes out.

It’s only as good as the “outstanding steps” it reads. If status lives in three places and is reconciled by hand days late, the tool nudges against a stale picture and chases students who already finished. Getting enrolment status into one current, trustworthy view — across the student management system, forms and inbox — is usually the first piece of work, and the piece that pays off across every intake after.

No. It removes the silent-stall detection and the repetitive “you’re nearly there” typing, so the officer spends their time on the students with real barriers — affordability, language, competing work. The judgement about who needs a phone call, a payment plan or a referral stays with them. The capacity it frees goes back into keeping students enrolled, not into the queue.

Current to the day. Census dates in Australian VET and higher education are hard deadlines, and a step marked outstanding might have been completed an hour ago. Nudge against a stale status and you chase the wrong people days before census. The honest test: can you say, right now, how fresh the outstanding-steps field is for a given student?

A student’s enrolment message, USI and identity documents are personal — and sometimes sensitive — information under the Privacy Act 1988 and the Australian Privacy Principles, and as an RTO you have obligations under the Student Identifiers Act for the USI. Any deployment runs against your own systems and data handling, not a shared pool — we scope where the data sits and who can see it as part of the build. The demo runs entirely on fabricated data; Aisha Rahman is not a real student.

It drafts to the steps and help paths you give it and flags the sensitive cases, but it does not certify compliance. Obligations under the Standards for RTOs as enforced by ASQA, subsidy rules like Skills First, and your own welfare and privacy policies are confirmed by a person. The tool gets the routine nudge most of the way; the accountable staff member closes the judgement calls.

What it takes to build

3–4 weeks · 4 phases

Reused from template~70%

Bespoke to this skin~30%

stack · Claude · enrolment-status retrieval · review UI

What it would cost

Fixed scope, fixed price, fixed dates.

Bite-sized first piece

One contained change, low risk

Pilot build

Most builds land here

Embedded support

Scale on proof

Considering this for your intake?

The honest place to start is a bite-sized first piece — one contained change, low risk. Tell us where students fall out of the funnel; we'll play it back, scope it, and show you what's possible.

Book a call How we work