Main site

Construction AI agent · Workforce currency

Workforce Licensing & Currency AI Agent

Licences, tickets and inductions expire on their own dates, not on yours. The agent tracks absence, currency and the clock, and never examines a document — the named employer representative does that, and attests.

Typical delivery
4–6 weeks
Deployment
Your stack
Officer attests
Per engagement
After launch
Agent Care

What this agent does

Tracks the currency of people, not the standing of firms

In

  1. 01

    A worker mobilises, and the agent opens a file against the role, the site and the jurisdiction over it.

  2. 02

    A role is scoped, and the agent separates what statute requires of it from what the client requires by contract.

Reason

  1. 03

    A US hire starts, and the agent runs the three-business-day Section 2 clock and reports the fields left blank.

  2. 04

    A UK engagement starts on or after 1 October 2026, and the agent tests whether a check exists before day one.

  3. 05

    A ticket falls due, and the agent counts the renewal window down against the work already rostered for that person.

Decide

  1. 06

    An Australian high risk task is rostered, and the agent checks the licence class held against the class it needs.

  2. 07

    A file is called for, and the agent produces what is held and names what is not, with where each item came from.

Out

  1. 08

    An engagement ends, and the retention clock runs on its own arithmetic in each jurisdiction, per person.

  2. 09

    Execute write actions only inside the approval boundaries agreed during implementation.

Product statement

The agent reports absence, currency and dates — never a view on a document, and never a list of people by nationality, visa type or country of issue, whoever is asking and whatever the reason given.

Example workflow

One worker, mobilisation to file

  1. 1Worker nominatedDirect labour, labour hire, or a subcontractor at any tier
  2. 2Requirements assembledThe role, the site, the jurisdiction and what each of them requires of this person
  3. 3File comparedWhat is held, what is absent, what expires and confidence
  4. 4Controls appliedAbsence checks, expiry-window checks, identity-match checks and confidence threshold

No human action required

Stages 1 to 4 run unaided, and nothing is examined, judged or attested at any of them — the agent compares records, and the representative's lane opens at the confidence gate.

5DecisionBranches at the confidence threshold
High confidence

Goes to the employer representative.

Low confidence

Adds a site manager read first.

Representative review

The file is held with what is absent, what expires and the confidence.

Clear · Hold · Send to site manager

Cleared — released to work

  1. 6Register and site access updatedOnly where write access and approval policy allow it
  2. 7Outcome evaluatedAbsences found late, expiries missed, holds overturned and files produced complete

Holds lifted

Every hold lifted in review is counted in the evaluation.

What should not run autonomously

Human approval stays in control

Outside the boundary — human approval required

8 items

  • Judging whether a document is genuine or belongs to the worker.
    • Signing or completing the attestation on the form.
    • Inferring, scoring or predicting immigration or citizenship status.
    • Listing workers by nationality, visa type or country of issue.

  • Automation boundary

    Agent acts unaided

    • ✓Map what the role, the site and the jurisdiction.
    • ✓Compare that against what the file already holds, item by item.
    • ✓Count down expiry, renewal and re-verification dates from the record.
    • ✓Report what is absent, and hold the file open.

    Any write happens inside the boundaries agreed at implementation, never ahead of an attestation.

    • Refusing or granting a worker entry to the site.
    • Certifying payroll or any statement about what was paid.
    • Granting, renewing or endorsing a licence, card or ticket.
  • Changes to currency rules or approval thresholds.

Example output

One worker, annotated

Subcontractor management checks the firm you engaged; this checks the person it sent.

Currency record · single worker

Illustrative example
Role
Mobile crane operator
Requirement
High risk work licence of the class the rostered lift actually needs
On file
Expired, in window
Currency
Renewable — retraining after
Confidence
91%
Consequence
The PCBU must see written evidence

As received

Taken from the register and the licence records as they stand — nothing on this side is written by the agent.

What was read

  • Employer licence register
  • Site induction record
  • High risk task roster

Why this reads amber

It lapsed inside the twelve-month renewal window, so it is renewable rather than gone.

What the score decides

Below the configured threshold the file picks up a site manager read before.

Action

ClearHoldSend to site manager

Value

Where AI adds value

The same four claims, placed at the point in the workflow where each one applies.

Where the value lands

Value 01 – 04

Every worker on file

From the register

03Currency

Check the person, not the firm

Read each requirement against the jurisdiction and the role on this site, and where the guidance is still draft, report it as draft.

01Approved path

It never reads a document

Routine renewals arrive already counted down against the work ahead.

02Human review

Send review to the ones that stop work

Absences and near expiries are marked, so the representative starts where the work is about to be blocked.

04

Build an evidence trail

The worker, the requirement it maps to and the officer who attested stay on the file.

Integrations

Typical integrations

Five system groups connect to the same agent. Which of them are in scope is decided in discovery.

  • Workforce and HR records

    • Workday
    • SuccessFactors
    • Employment Hero
    • BambooHR

  • Site access and induction

    • Damstra
    • Rapid Global
    • Biosite
    • SALTO site access

  • Licence and card registers

    • CSCS Smart Check
    • CITB
    • State WHS licence registers

Agent

Workforce licensing and currency

  • Reads the register
  • Assembles the file
  • Holds for the officer

  • Verification and inspection

    • E-Verify
    • Home Office check
    • Notice of Inspection pack

  • Observability & evaluation

    • OpenTelemetry
    • Langfuse
    • Supported monitoring/evaluation sources

Integration availability depends on the client's existing systems and API access.

Agent controls

Six layers between the model and the attestation

The six narrow as they run. What none of them stops appears in the map beneath it.

L6 · Outermost — last line of defence

Inward → L1 · closest to the model

L6Rollback / safe modePull the agent back to reporting only when evaluation or production signals degrade.Roll back
L5Version monitoringTrack model, prompt, currency-rule and jurisdiction configuration changes.Track
L4TraceabilityRecord what was read, what was reported absent, and who attested when.Record
L3Representative gateHold each file for the named representative; it gates release, not whether the record is complete.Gate
L2Policy guardrailsTest each output against the rules barring status inference and document judgement; a failure returns it.Restrict
L1Confidence thresholdsRoute low-confidence files to a site manager read first.Require review
Model coreFile assembled — requirements, absences, expiry dates and confidence
L1 – L2
Test whether a file may stand
L3
Puts the release in a representative's hands
L4 – L5
Keep the worker and the requirement behind it
L6
Holds the file at incomplete when signals degrade

How Nestack evaluates it

Evaluate the currency workflow — not only the finished file.

Coverage runs the whole depth of the workflow, and every layer is cut by slice.

Surface — the file an inspector pulls

  1. E1Final-output evaluationDid the file name every requirement that role and site carry?
  2. E2Step-level evaluationDid the agent read the jurisdiction and the site rules in force?
  3. E3Tool evaluationDid it read and write the correct person and the correct register?
  4. E4Confidence calibrationDo low-confidence files actually attract more holds?
  5. E5Slice evaluationHow does performance change across specific engagement types?
  6. E6Business outcomeHow many absences surfaced only after the person was already working?

Floor — the record the representative attests

Failure modes

Where each failure originates in the agent

Seven failure modes, each set at the stage where it first appears.

Agent lifecycle

Direction of processing →

  1. 01 · Retrieval

    1 mode

    • GI-01

      Register misses the chain

      Labour hire and lower tiers are not in the payroll source.

    Stage gathers

    Registers, licence records, inductions and rosters

  2. 02 · Reasoning

    2 modes

    • GI-02

      Licence class not tested

      A licence exists and is current, and is the wrong class.

    • GI-03

      Renewal window read as gone

      An expired licence still inside its window reads as lost.

    Stage proposes

    Requirements, absences, dates and confidence

  3. 03 · Tool / write

    2 modes

    • GI-04

      Two workers, one name

      Credentials attach to the wrong person of that name.

    • GI-05

      Clock never started

      The hire date is right and the three-day clock is not run.

    Stage writes

    Only where write access and approval policy allow it

  4. 04 · Output

    1 mode

    • GI-06

      Completeness read as fine

      Form fields are full and the audit trail was never checked.

    Stage returns

    The file the representative attests and an auditor pulls

  5. 05 · Change / Version

    1 mode

    • GI-07

      Start date inherited

      A new engagement carries the old date past 1 October.

    Stage tracks

    Model, prompt, currency rules and site config

  • Sev-1 · the agent performs a held act
  • Sev-2 · an absence reaches the roster
  • Sev-3 · a register degrades, file to review
See the complete Construction failure-mode catalogue

Affected slices

A clean site figure can rest on one trade

A site-level file-completeness figure can read well while a few trades and engagement types carry the missing items. No regulator publishes a construction breakout, so Nestack reports yours by slice.

Slice performance — reported separately, not only in aggregate

Illustrative example
SliceFailure rateLiftLift vs. thresholdStatus
Third-tier subcontract9.1%3.7×Review
Labour hire and agency6.9%2.8×Review
High risk licensed trades4.0%1.6×Watch
Direct labour, one site1.9%0.8×Normal

Bar: missing-item-rate lift vs. a direct-labour baseline · scale 0–4.0× · tick marks the 2.0× review threshold

2 of 4 slices over threshold

Evidence-linked improvement

Every cycle leaves one more test behind

A cycle closes when the lapsed ticket is a regression case. That suite is what the next start worked is measured against.

Improvement cycle · five stages

Switchback — the path turns at Improve and returns at Learn

  1. 01

    Detect

    Missing-item rate rises in one trade.

  2. 02

    Diagnose

    The ticket that expired between induction and the lift it was needed for is traced back to one step.

  3. 03

    Improve

    The fix carries a number, and the sites that forced it travel with it.

  4. 04

    Verify

    Nothing goes out until every touched worker case has passed again.

  5. 05

    Learn

    It is kept permanently, and the currency rules are amended alongside it.

  6. Learn → Detect

    The return edge. The next detection runs against a suite one case longer.

Typical build scope

Twelve workstreams across six weeks

The build scope read against the delivery timeline. Week structure follows the six-week plan — discovery, sources, currency workflow, evaluation, integration, then production validation and handover.

Workstream

Week 1Week 2Week 3Week 4Week 5Week 6

  1. 01Workforce currency discovery and boundary definition.Week 1
  2. 02Register and HR source assessment.Week 1, Week 2
  3. 03Jurisdiction, role and site requirement mapping.Week 1, Week 2
  4. 04Person register ingestion and matching.Week 2
  5. 05Requirement comparison and absence logic.Week 2, Week 3
  6. 06Confidence scoring and file routing.Week 3
  7. 07Employer representative review workflow.Week 3, Week 4
  8. 08Register and site-access integration.Week 5
  9. 09Absence and expiry cases.Week 4, Week 5
  10. 10Guardrails and attestation controls.Week 4
  11. 11Worker-trail instrumentation.Week 5, Week 6
  12. 12Deployment, documentation and Agent Care handover.Week 6

12 workstreams · 6 weeks · bar shows the weeks a workstream is active — several run in parallel

Final scope and sequence confirmed in discovery

Engagement tiers

What each tier includes

Rows are the capabilities named in each tier's scope. Higher tiers include everything below them.

Capability✓ in scope · — not at this tierPilotOne site, one tradeProductionProduction register systemsAdvancedMultiple sites / jurisdictions
Introduced at Pilot
Currency checks to your rules✓✓✓
Representative review gate✓✓✓
File-completeness baseline✓✓✓
Introduced at Production
Reporting by trade—✓✓
Review workflow in your systems—✓✓
Approved write-back—✓✓
Site-access integration—✓✓
Introduced at Advanced
Multi-jurisdiction rule sets——✓
Multi-stage site approvals——✓
High headcount——✓
Multi-site currency controls——✓
Build priceFrom $5,000From $8,000Custom quote
Final build price
Confirmed after discovery based on integrations, workflow complexity, workforce size, approval controls and deployment requirements.
Separate from build
Build pricing is separate from recurring Agent Care, which covers managed monitoring, evaluations, incidents and verified improvements after launch.

What we need from you

What you bring, and what we build with it

Each input maps to a piece of build scope and a week in the delivery timeline.

  1. 01Your worker register and engagement types

    Person-level register ingestion and matching

    Week 1

  2. 02Representative roles and the sites they cover

    Requirement mapping and the absence-detection baseline

    Week 2

  3. 03Your jurisdiction, role and site requirement rules

    Requirement-rule mapping and automation-boundary definition

    Week 1

  4. 04Access to relevant APIs, feeds or exports

    Register, HR and site-access assessment, then integration setup

    Week 2

  5. 05Files you would not want inspected

    Expiry cases and the evaluation suite

    Week 4

  6. 06What no attestation may borrow

    Confidence scoring, file routing, guardrails and approval controls

    Week 3

  7. 07Named employer representatives to review files

    Representative review workflow, then pilot and production validation

    Weeks 5–6

Nothing else is required

Deployment, documentation and Agent Care handover are ours.

Delivery timeline

Four phases across six weeks

These widths track the weeks the work genuinely fills, so the fifth band is doubled on purpose.

Phase

W1

W2

W3

W4

W5

W6

  1. DiscoveryW1
  2. BuildW2 – W3
  3. EvaluateW4 – W5
  4. Pilot & LaunchW5 – W6

Week focus

W1Workforce currency discovery, rule mapping and the automation boundary

W2Register integration and the absence-detection baseline

W3Requirement comparison, confidence logic and review controls

W4Evaluation suite, guardrails and failure-mode testing

W5Site-access integration, pilot crews and targeted corrections

W6One mobilisation run under the employer representative, then Agent Care handover

Reading the band
A bar spans only the weeks its own work is named in. The week 5 overlap is worked time, not padding.
At the end of W6
The final checks clear on live crews and monitoring moves to Agent Care.
Duration
Six-week plan shown · typical delivery 4–6 weeks depending on scope confirmed in discovery.

Next step · Construction AI agent

Build a workforce currency agent around the attestation you already carry.

Three jurisdictions, three accountable acts: a perjury signature in the United States, no signature at all in the United Kingdom where the absent check is the liability, and an Australian duty to see the evidence first. Never a view on anybody status — acting on such a flag is itself the offence.

Typical delivery
4–6 weeks
Deployment
Your registers
Officer attests
Per engagement
After launch
Agent Care available

Nestack Agents · Workforce licensing and currency

AGT-CN-12 · Agent Care available after launch

Build a workforce currency agent

Four questions so the call starts from your actual setup — then pick a time that suits you.

About you
Your workforce

No obligation. We’ll map the workflow, name the automation boundary and tell you if an agent is the wrong answer.