Mobilisation Pack Drafter | Real Minds AI
Construction /Document Generation interactive demo field guide · 10 min

Mobilisation Pack Drafter

Reads the sales handoff on a won job, splits it into one job card per crew leg — scope, treatment, spec, quantities, site constraints — and lays the pack out for the works manager to check and approve before any crew is booked.

theater/demos/construction_mobilisation-pack-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 won job's handoff, splits it into one card per crew leg with the scope, treatment, spec and site constraints each crew actually needs, and lays the pack out for the works manager to approve before anything is booked.

Input 01
The won job, as sales handed it over

The accepted quote or works order — the stages, the dates and time windows, the crew against each stage, the treatments and specifications, the quantities, and the site access conditions written into the job.

Agent 02
Splits it by crew and by day

Pulls out each stage as a separate leg, carries the treatment, specification and quantities onto it, keeps the order the stages depend on, and checks each leg's proposed hours against the access conditions the job carries.

Output 03
A pack, with the awkward one flagged

One job card per crew with everything a supervisor needs on site, a calendar slot for each, and an amber flag on any leg whose hours clash with a site restriction — laid out for the works manager to approve, edit or reject before a single crew is booked.

Where it works well

It does the whole ritual every time, and the ritual is what the volume is made of.

  • Best where jobs are small and frequent rather than long and few — a crew on several jobs a week, each carrying the full delivery ritual at small scale, is where the per-job overhead multiplies.
  • Every card carries the same fields in the same places, so a supervisor reads the same sheet on every job instead of decoding a different whiteboard photo each morning.
  • The recaptured time goes back to the works manager for the judgement calls — which crew, what sequence, whether the site is really ready — not the transcription underneath them.

A surfacing job is not one job. It is a spray seal crew on Tuesday, an asphalt crew on Wednesday and a line marking crew on Thursday, each needing its own scope, its own treatment, its own specification and its own site conditions. Doing that once is twenty minutes at a whiteboard. The cost is doing it again on Monday, and again on Wednesday, and again the week after.

Where it works badly

It only knows what the handoff says — and a handoff that leaves things out still produces a confident, tidy pack.

  • It does not know who is actually free. It reads the dates the job carries; it has no view of the crew already committed to another site that Tuesday, so it will happily draft a card that cannot be worked.
  • It does not know the weather, and spray sealing is weather-dependent — a seal that cannot be laid wet or below the binder's temperature floor is a decision for the person reading the forecast, not for a document reader.
  • If your jobs are won by phone and programmed straight onto a whiteboard with nothing written down in between, there is no handoff to read, and getting one written is the real work before any of this helps.
The honest test

Pick last week's most awkward job. If the sequencing, the site access restriction and the specification were all written down somewhere before the crew left the yard — this helps. If they were in someone's head, this makes the gap faster to reach, not smaller.

Give it a job where the sequencing is understood rather than written, or where the site restriction lives in the estimator's head, and it drafts three clean cards with the order wrong or the constraint missing. The pack looks identical whether the handoff was complete or not. That is the trap.

What it doesn't do — and shouldn't

It drafts the pack. The works manager books the crews. That line is deliberate.

WHAT IT DOES
Lays out every leg with its scope, treatment, specification and quantities
Carries the site access conditions the job states onto every affected card
Flags a leg whose proposed hours clash with a restriction the job carries
WHAT IT WON’T
Book a crew or send anything to a supervisor
Decide the sequence, or move a leg to a different day
Confirm the crew is free, the site is ready, or the weather will hold

Mobilising a crew is the moment the job stops being paperwork. A crew sent with the wrong treatment lays the wrong surface; a crew sent into a school zone in the wrong window is a traffic management failure. Those sit under the Work Health and Safety Act 2020 (WA) and its General Regulations, and under the temporary traffic management frame Main Roads works to — AS 1742.3, the Austroads Guide to Temporary Traffic Management, and the Traffic Management for Works on Roads Code of Practice. The consequence lands on the business and on the crew, so the manager stays on the decision.

What your data has to look like

A written handoff on every won job, with the sequencing and the site conditions actually in it.

32%
Typical readiness
across orgs we see, before the first job
A written handoff per won job
Needs shaping
Stage sequencing stated, not assumed
Needs shaping
Site access conditions captured at quote time
Needs shaping
Consistent crew names
Usual weak point
Treatments and specifications written as references
Usual weak point
The real first job

The handoff itself is usually the weak point. In a lot of surfacing businesses the job goes from won to programmed with nothing written in between, because everyone involved already knows. Getting that handoff into a consistent written form — including the sequencing and the site conditions — is normally the real first piece of work, and it is worth doing whether or not anything is ever drafted on top of it.

Right fit if…
Your crews run several jobs a week, each with its own quote, specification and site conditions
A won job routinely splits across two or more crews on different days
Someone programmes the work onto a whiteboard or a spreadsheet from a written quote
Your site constraints — zone windows, plant hours, traffic management — are written on the job today
Walk away if…
You run a small number of long projects rather than many short ones
Jobs go from won to programmed by conversation, with nothing written in between
The sequencing and site conditions live with one experienced person rather than on the job
You want the tool to book crews or commit the programme without a manager approving it
Open questions

The worried-buyer questions, answered straight

That is exactly why nothing reaches a crew on its say-so. Every card shows the treatment, the specification and the quantities it took from the job, so the works manager is checking a draft against the quote rather than typing one from scratch. It also flags the legs it is unsure about rather than smoothing them over — a leg whose hours clash with a site restriction comes back amber, not green. The manager approves; the pack goes nowhere until they do.
Then it will draft a tidy pack with those conditions missing, which is the honest failure mode. It reads what the handoff says; it cannot recover a restriction that was never written down. Getting the site conditions and the stage sequencing onto the job at quote time is usually the first piece of work, and it is worth doing on its own — it is the same gap that bites when the person who knew is on leave.
No. It removes the transcription — copying the same scope, treatment, specification and site conditions onto a card for each crew, several times a week. What is left is the part that was always the job: which crew, what order, whether the site is actually ready, whether the weather will hold. The programme is still theirs. The time it gives back goes into those calls, and into the jobs that would otherwise wait.
Current to the job as accepted. A superseded specification reference or a traffic management plan that has since been revised will be carried onto the cards exactly as the handoff states it, because that is all the tool can see. The honest test: is the version of the job it is reading the version the crew will actually work to? If quotes get revised after acceptance and the revision does not flow back, the pack is drafted from the wrong one.
Your quotes, rates, crews and client sites are commercially sensitive. Any deployment runs against your own systems and your own data handling, not a shared pool, and where the data sits and who can reach it is scoped as part of the build. The demo on this page runs entirely on fabricated data — the Main Street resurfacing job, its crews and its contact are invented, and no client work is shown.
No, and it is not trying to. It takes the dates and crews the job already carries and lays them out properly; it does not schedule, does not know who is free, and does not resolve a clash between two jobs competing for the same crew. Sequencing and allocation are judgement calls with real consequences on a site, and they stay with the person who owns the programme.
What it takes to build
3–5 weeks · 4 phases
Reused from template~70%
Bespoke to this skin~30%
stack · Claude · document parser · card renderer · review UI
What it would cost

Fixed scope, fixed price, fixed dates.

01
Bite-sized first piece
One contained change, low risk
02
Pilot build
Most builds land here
03
Embedded support
Scale on proof

Programming work off a whiteboard every week?

The honest place to start is one job type, one crew mix, low risk. Tell us what happens between winning the job and the crew leaving the yard; we'll play it back, scope it, and show you what's possible.

More in Construction
Tender Compliance Matrix Builder
View →
Timesheet-to-Payroll Reconciler
View →
Supplier-Invoice Reconciler
View →
RFI Triage Desk
View →
Ask us anything