Main site

Mining AI agent · Tenure & permits

Tenement & Permit-Tracking AI Agent

Track what each tenement and permit obliges you to do — expiry, expenditure, rent, reporting and the conditions on the grant — assemble the evidence, and leave every lodgement to the tenement manager.

Typical delivery
4–6 weeks
Deployment
Your stack
Lodgement
Named person
After launch
Agent Care

What this agent does

Tracks the obligation, does not discharge it

In

  1. 01

    Take the tenement and permit register, the grant documents and the conditions written on each title.

  2. 02

    Read the expenditure ledger, rent and rates notices, reporting deadlines and the clearances already held.

Reason

  1. 03

    Date every obligation on a title — expiry, renewal, rent, minimum expenditure, reporting and relinquishment.

  2. 04

    Match spend, reports, payments and approvals to the obligation each one discharges and the title it sits on.

  3. 05

    Read the registry's own record back against the company's and keep the fields where they disagree, dated to the extract.

Decide

  1. 06

    Flag a commitment running short, a date in reach and a condition with no evidence attached to it yet.

  2. 07

    Route heritage, native-title and landholder matters to the people and the counsel who own them.

Out

  1. 08

    Present the obligation pack — the date, the condition as written, the evidence found and what is missing.

  2. 09

    Retain the sources read, the register version, the match made, the confidence and every manager amendment.

Product statement

The agent tracks, matches and assembles. Lodging, renewing, surrendering, varying a permit and deciding a condition has been met stay with the tenement manager.

Example workflow

One title's year, end to end

  1. 1Titles drawnThe tenement and permit register, each grant document and the conditions endorsed on it
  2. 2Obligations datedExpiry, renewal, rent, minimum expenditure, reporting and relinquishment, on that jurisdiction's rules
  3. 3Evidence matchedSpend, reports, payments, approvals and clearances matched to the obligation each one discharges
  4. 4Records reconciledThe registry's record read against the company's, field by field, with each difference dated to its extract

No human action required

Stages 1 to 4 run without a person in the loop — dating, matching and the reconciliation finish before anyone is asked to read anything. Nothing is lodged in that stretch.

5DecisionSplits on evidence confidence and whether heritage, native title or a landholder is involved
Dated, evidenced, uncontested

Reaches the tenement manager as a complete pack.

Evidence missing, or heritage involved

Goes to the person who owns that obligation first.

Tenement manager

Reads the obligation, the condition as written, the evidence attached and what is missing, then decides what is lodged and when.

Accept pack · Amend · Refer

Pack accepted — handed back

  1. 6Pack handed overHanded to the tenement manager as a draft; nothing is lodged, renewed, surrendered or varied by the agent
  2. 7Outcome evaluatedWhat the manager amended, what the registry queried, and what fell due with no evidence behind it

Amendments

What the manager re-dates or refers is counted in the evaluation.

What should not run autonomously

Human approval stays in control

Outside the boundary — human approval required

8 items

  • Lodging, renewing, surrendering or transferring a tenement.
    • Applying for or varying a permit or licence.
    • Deciding that a condition has been met.
    • Determining a heritage, native-title or landholder obligation.

  • Automation boundary

    Agent acts unaided

    • ✓Date each obligation from the grant's own conditions.
    • ✓Match spend, reports, payments and clearances to the obligation each discharges.
    • ✓Read the registry's record back against the company's and keep the differences it finds, dated to the extract.
    • ✓Assemble the pack, naming the evidence still missing.

    Write actions run only inside the approval boundaries agreed during implementation. A lodgement is never one of them.

    • Certifying an expenditure or operations report.
    • Nominating the ground given up at relinquishment.
    • Releasing a rent, rates or security payment.
  • Giving legal advice on what a condition requires.

Example output

One obligation on one title, annotated

Everything the agent proposes is attached to the title and the obligation it came from.

Obligation pack · one title, one obligation

Illustrative example
Title
Exploration licence
Obligation
Minimum annual expenditure
Due in
34 days
Proposed
Raise as a shortfall
Confidence
90%
Condition met
Not decided by the agent

As received

The title, the obligation in the grant's own words and the date it falls due — read from the register, not restated.

Evidence used

  • Expenditure ledger lines
  • Registry record, this week
  • Condition text on the grant

Why a shortfall

Part of the spend claimed to this title was incurred on ground since given up.

What the score decides

Confidence decides how hard the manager reads this, not whether the commitment is met.

Action

Accept packAmendRefer

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 title you hold

From the tenement and permit register

03Obligations

Work from the words on the grant

Use the conditions endorsed on each title, the clocks that jurisdiction runs, and the agreements already in place.

01Approved path

Keep the quiet obligations visible

A rent notice, a report and a relinquishment date carry the same forfeiture risk as a headline commitment, and they are the ones that go unwatched.

02Human review

Reach it while it can still be met

A commitment running short, or a clearance not yet held, reaches the tenement manager with time to act rather than after the date has passed.

04

Build an evidence trail

Retain the sources read, the register version, the evidence matched to each obligation, the confidence and every amendment — on both paths.

Integrations

Typical integrations

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

  • Tenement systems

    • Landtrack
    • Global Tenements
    • Trimble
    • In-house tenement registers

  • Registries & approvals

    • State portals
    • registry extracts
    • Approvals registers
    • gazettal notices

  • Finance & expenditure

    • SAP
    • Pronto
    • Project codes
    • rent and rates invoices

Agent

Tenement & permit tracking

  • Dates each obligation
  • Matches the evidence
  • Routes what is missing

  • Agreements & land access

    • Heritage and native-title agreements
    • Landholder access and compensation

  • 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 tenement file

Each control wraps the one inside it. An obligation clears every layer before the tenement manager reads it, and the lodgement sits outside all six.

L6 · Outermost — last line of defence

Inward → L1 · closest to the model

L6Rollback / safe modeReturn tracking to the tenement team if evaluations or production signals degrade.Roll back
L5Rule-set versioningTrack calendar, rule-set, condition-text, model and prompt changes, and which version dated an obligation.Track
L4Change controlRecord the sources read, the register version, the proposal, the named approver and every amendment.Record
L3Owner gateLodgement, condition sign-off and any heritage or native-title question stay with the named owner.Gate
L2Evidence matchSpend, reports and payments are matched to the obligation they discharge; what does not match is flagged, not counted.Flag
L1Source and rule setEach dated obligation carries the grant or condition it was read from and the rule set that dated it.Cite
Model coreObligation proposed — the date, the condition as written, the evidence matched and confidence
L1 – L2
Decide whether a dated obligation may stand
L3
Decides who may lodge and who may sign
L4 – L5
Keep the record and the rule sets honest
L6
Pulls automation back when signals degrade

How Nestack evaluates it

Evaluate the whole obligation — not only the pack the manager opens.

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

Surface — the obligation pack the tenement manager opens

  1. E1Final-output evaluationWas the obligation dated correctly for that title and jurisdiction?
  2. E2Step-level evaluationDid it read the current grant, condition text and rule set?
  3. E3Tool evaluationDid it resolve the correct title, registry record and cost code?
  4. E4Evidence calibrationDoes what was matched actually answer the obligation claimed?
  5. E5Slice evaluationHow does tracking quality change across specific title cohorts?
  6. E6Business outcomeWhat fell due unevidenced, was queried by the registry or was re-dated?

Floor — the ground the company still holds

Failure modes

Where each failure originates in the agent

Seven failure modes plotted against the five stages of the agent lifecycle.

Agent lifecycle

Direction of processing →

  1. 01 · Retrieval

    2 modes

    • TP-01

      Superseded grant read

      A varied condition is dated from the old grant.

    • TP-02

      Stale registry extract

      The pack is built from an extract taken months ago.

    Stage gathers

    Grants, conditions, registry records and spend lines

  2. 02 · Dating

    2 modes

    • TP-03

      Clock started on the wrong date

      A renewal or reporting date is set by the wrong rule.

    • TP-04

      Spend counted to the wrong ground

      Expenditure on relinquished ground is claimed to the title.

    Stage proposes

    Obligation dates, amounts due and the rule set applied

  3. 03 · Evidence match

    1 mode

    • TP-05

      Evidence does not discharge

      A report is matched to a condition it does not answer.

    Stage checks

    What discharges an obligation and what is still missing

  4. 04 · Handoff / write

    1 mode

    • TP-06

      Heritage duty reduced to a date

      A clearance date is presented as the obligation itself.

    Stage presents

    The obligation pack the manager reads and its queue

  5. 05 · Change / Version

    1 mode

    • TP-07

      Rule-set change re-dates quietly

      A calendar change moves dates with no named approver.

    Stage tracks

    Model, prompt, rule-set and condition-text changes

  • Sev-1 · ground or a legal duty is put at risk
  • Sev-2 · a wrong obligation reaches the manager
  • Sev-3 · tracking goes stale, more is re-dated
See the complete Mining failure-mode catalogue

Affected slices

The obligations that get missed sit on the newest ground

Exposure follows how recently ground arrived. A decade-old lease repeats its obligations yearly; ground from an acquisition, second-jurisdiction titles and surrender years do not. Nestack reports performance by slice, not only in total.

Slice performance — reported separately, not only in aggregate

Illustrative example
SliceFailure rateLiftLift vs. thresholdStatus
Newly acquired tenements5.7%3.9×Review
Titles in a second jurisdiction4.1%2.8×Review
Relinquishment and surrender years2.7%1.8×Watch
Long-held granted mining leases1.3%0.9×Normal

Bar: re-dated-or-missed obligation rate, lift vs. long-held lease baseline · scale 0–4.0× · tick marks the 2.0× review threshold

2 of 4 slices over threshold

Evidence-linked improvement

A missed date is a wrong rule, not a diary slip

The fix is not a date corrected in a calendar. What changes is the rule that dated it, the test the evidence is held to, or the source the obligation was read from.

Improvement cycle · five stages

Switchback — the path turns at Improve and returns at Learn

  1. 01

    Detect

    Re-dated obligations or registry queries rise on a title cohort.

  2. 02

    Diagnose

    The rule that produced the date is read before the date itself is touched.

  3. 03

    Improve

    A named owner approves the corrected rule, dated from the period it applies.

  4. 04

    Verify

    Every title on that rule set is dated again and read against what happened.

  5. 05

    Learn

    The corrected rule is written into the register the next year is dated from.

  6. Learn → Detect

    The return edge. A rule that dates an obligation is changed by a named owner, before the next cycle runs on it.

Typical build scope

Twelve workstreams across six weeks

The build scope read against the delivery timeline. Week structure follows the six-week plan — discovery, titles and rule sets, dating and evidence logic, evaluation, integration, then production validation and handover.

Workstream

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

  1. 01Workflow discovery and automation-boundary definition.Week 1
  2. 02Tenement and permit register assessment.Week 1, Week 2
  3. 03Jurisdiction rule sets and obligation calendars.Week 1, Week 2
  4. 04Grant and condition-text extraction.Week 2
  5. 05Expenditure, rent and reporting data access.Week 2, Week 3
  6. 06Evidence matching and obligation tests.Week 3
  7. 07Registry reconciliation and differences.Week 3, Week 4
  8. 08Agreement and heritage obligation routing.Week 4
  9. 09Evaluation suite, date recall and regression cases.Week 4, Week 5
  10. 10Manager review and amendment capture.Week 5
  11. 11Tenement-system integration and write-back.Week 5, Week 6
  12. 12Observability, deployment 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 jurisdiction, one registerProductionProduction register integrationAdvancedMulti-jurisdiction portfolio
Introduced at Pilot
Obligation calendar dated from the grant✓✓✓
Condition text kept beside each obligation✓✓✓
Evidence matched to each obligation✓✓✓
Registry and company record reconciliation✓✓✓
Heritage and native-title matters routed to owners✓✓✓
Lodgement stays with the tenement manager✓✓✓
Baseline evaluation✓✓✓
Introduced at Production
Manager review workflow and write-back—✓✓
Observability and evaluation—✓✓
Introduced at Advanced
Multiple jurisdictions and rule sets——✓
Group portfolios and enterprise controls——✓
Build priceFrom $5,000From $8,000Custom quote
Final build price
Confirmed after discovery based on the number of titles and jurisdictions, register and finance integrations, condition and agreement complexity, 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 tenement and permit register as it actually stands

    Tenement and permit register assessment

    Week 1

  2. 02The jurisdictions you hold ground in and their calendars

    Jurisdiction rule sets and obligation calendars

    Week 1

  3. 03Grant documents and the conditions endorsed on each title

    Grant and condition-text extraction

    Week 2

  4. 04Access to expenditure, rent and reporting records

    Expenditure, rent and reporting data access

    Week 2

  5. 05Your heritage, native-title and landholder agreements, and who owns them

    Agreement and heritage obligation routing

    Week 4

  6. 06Obligations you have missed, and the ones you nearly did

    Evaluation suite, date recall and regression cases

    Week 4

  7. 07A named tenement manager to review packs

    Manager review workflow, then pilot packs 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

Phases are drawn over the weeks they actually occupy. Week 5 carries both the evaluation work and the first packs a tenement manager reads.

Phase

W1

W2

W3

W4

W5

W6

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

Week focus

W1Register, jurisdictions and the automation boundary

W2Grant documents, condition text and the expenditure data wired in

W3Obligation dating and evidence matching on the first jurisdiction

W4Evaluation suite, date recall and registry reconciliation

W5Register write-back, supervised packs and targeted corrections

W6Managers work live packs beside your own calendar, then Agent Care starts

Reading the band
Week 4 re-dates title years that have already run, so the dating rules are argued out against real outcomes before a live year depends on them.
At the end of W6
The manager who lodges has worked a live obligation cycle beside the existing calendar, then Agent Care takes over monitoring.
Duration
Six-week plan shown · typical delivery 4–6 weeks depending on scope confirmed in discovery.

Next step · Mining AI agent

Build a tenement-tracking agent around your obligation calendar.

Show us your tenement register, the grant documents behind it and who lodges what today. We'll date one title's year from its conditions and put the rule behind each date beside it, so you can argue with the rule.

Typical delivery
4–6 weeks
Deployment
Your tenement stack
Lodgement
Tenement manager, always
After launch
Agent Care available

Nestack Agents · Tenement & permit tracking

AGT-MN-04 · Agent Care available after launch

Build a tenement-tracking agent

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

About you
Your setup

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