Nestack Agent Care
Industries / Accounting / Lease & fixed-asset agent

Accounting AI agent · Lease & fixed assets

Lease & Fixed-Asset Accounting AI Agent (ASC 842 / IFRS 16)

Extract lease terms from the agreement, propose the classification and discount rate, build the right-of-use and liability schedules, maintain the asset register, and flag remeasurement triggers, with every posting accepted by the financial controller.

4–6 weeksTypical delivery
Your stackDeployment
Pre-postingController approval
Agent CareAfter launch

What this agent does

Extracts the terms, never the conclusion

In
01

A lease or service agreement arrives from the contract repository, ERP or lease-accounting system.

02

A term, payment schedule and option set are normalised, each figure keeping the clause it came from.

Reason
03

An agreement is screened for an identified asset and a right to control its use, surfacing embedded-lease candidates.

04

A classification test and a proposed discount-rate input are drafted against the configured rate hierarchy.

05

A right-of-use and liability schedule is built, and the fixed-asset register carries placed-in-service dates forward.

Decide
06

A renewal, option exercise or contingency resolution is checked against the modification-versus-remeasurement rule.

07

An embedded-lease candidate, a classification or a rate proposal is routed to the named lease accountant for review.

Out
08

A clause, a schedule, its flags and the accountant's disposition are retained against the lease and the register.

09

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

Product statement

The agent proposes the classification and the rate input; the financial controller decides, posts every entry, and the company stays the preparer of record.

Example workflow

One lease, agreement to posting

AgentHuman
1Agreement receivedLease agreement, amendment, contract repository entry or ERP record
2Terms assembledPayment schedule, options, embedded-lease candidates and prior schedule, each with its clause
3Schedule proposedClassification, rate input, schedule and confidence
4Controls appliedIdentified-asset check, classification-test check, remeasurement gate and confidence threshold
No human action required

Stages 1 to 4 run unaided, and nothing posts at any of them — the agent is drafting the schedule, and the accountant's lane opens at the confidence gate.

5DecisionBranches at the confidence threshold
High confidence

Goes to the lease accountant to approve.

Low confidence

Adds a senior-accountant read first.

Controller approval

The schedule is held with its source clauses, its flagged triggers and the confidence.

Approve · Edit · Send to senior review
Approved — ready to post
6Asset register updatedOnly where write access and approval policy allow it
7Outcome evaluatedClassification correctness, remeasurement outcomes, controller edits and post-posting corrections
Edits

Every controller edit is counted in the evaluation.

What should not run autonomously

Human approval stays in control

Outside the boundary — human approval required8 items
Concluding that a contract contains a lease.
Setting the discount rate.
Classifying a lease as finance or operating.
Recording an impairment.
Automation boundaryAgent acts unaided
Extract lease terms and flag embedded-lease candidates for review.
Apply the configured classification tests and the rate hierarchy.
Build the schedules, and roll the fixed-asset register forward.
Flag remeasurement triggers and hold the schedule.
Any write happens inside the boundaries agreed at implementation, never ahead of approval.
Deciding a useful life or a placed-in-service date.
Confirming a common-control related-party relationship.
Electing a short-term or low-value class exemption.
Posting an entry to the register or the ledger.

Example output

One schedule line, annotated

Everything the agent builds is attached to the clause it was drawn from.

Lease-schedule output · single leaseIllustrative example
Lease
Schedule line
Liability (PV)
Source clause
Confidence
Preparer of record
Distribution-center lease
Ten-year term with two five-year renewal options, proposed at the incremental borrowing rate
$1,240,000
Lease agreement, Section 4
89%
Lease accountant name and title
As receivedTaken from the lease agreement and the prior schedule — nothing on this side is a conclusion the agent reached.
Source clauses used Lease agreement clause Renewal-option clause Prior-period schedule
Why this inputThe agreement states the term and the options; whether the rate resets.
ActionApproveEditSend to senior review
What the score decidesBelow the configured threshold the schedule picks up a senior read before it reaches.

Value

Where AI adds value

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

Where the value landsValue 01 – 04
Every leaseFrom the agreement
03Scheduling

Draft from the agreement

Draw on the clauses on file and the firm's configured classification and rate rules.

01Approved path

Not every change is a modification

Routine schedules and register entries arrive already built.

02Human review

Send review to the risky leases

Embedded-lease candidates and remeasurement flags are marked, so the accountant's read starts where risk concentrates.

04Build an evidence trail

The term, the clause it was drawn from and the accountant who booked it stay on the schedule.

Integrations

Typical integrations

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

Lease-accounting systemsLeaseQuery · Visual Lease
CoStar Real Estate Manager
ERP and general ledgerNetSuite · SAP
Oracle · Microsoft Dynamics
Document repositoryContract-management systems
Shared drives · DMS platforms

Agent

Lease & fixed-asset accounting

Reads the agreement
Builds the schedule
Holds for approval

Fixed-asset registerSage Fixed Assets
Bloomberg Tax Fixed Assets
Observability & evaluationOpenTelemetry · 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 posted schedule

The controls sit one inside the next. What they let through is named in the map below.

L6 · Outermost — last line of defenceInward → L1 · closest to the model
L6Rollback / safe modePull the agent back to term extraction only when evaluation or production signals degrade.Roll back
L5Version monitoringTrack model, prompt, classification-rule and rate-configuration changes.Track
L4Schedule trailRecord the source clause, the schedule, the flags and the accountant's disposition.Record
L3Controller approvalHold the schedule for the named accountant; it governs posting, not whether the classification is right.Gate
L2Policy guardrailsTest the schedule against configured classification tests and the rate hierarchy; a failure returns it.Restrict
L1Confidence thresholdsRoute low-confidence schedules to a senior read before the accountant sees them.Require review
Model coreSchedule produced — classification, rate input, ROU and liability figures and confidence
L1 – L2Test whether a schedule may stand
L3Puts the posting in the accountant's hands
L4 – L5Keep the term and the clause behind it
L6Reverts to term extraction only when signals degrade

How Nestack evaluates it

Evaluate the drafting and posting workflow — not only the final schedule.

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

Surface — the schedule the controller signs
Depth of coverage ▼
E1Final-output evaluationDid every extracted term match its source clause?
E2Step-level evaluationDid the agent use the right classification tests and rate hierarchy?
E3Tool evaluationDid it read and write the correct lease and the correct register entry?
E4Confidence calibrationDo low-confidence schedules actually attract more accountant edits?
E5Slice evaluationHow does performance change across specific asset classes?
E6Business outcomeHow many schedules needed an accountant edit or a correction after posting?
Floor — the outcome the controller answers for

Failure modes

Where each failure originates in the agent

Seven failure modes, placed at the stage each one originates.

Agent lifecycleDirection of processing →
01 · Retrieval1 mode
RO-03

Superseded amendment used

The schedule is built from an earlier draft of the lease, not the fully executed version.

Stage gathersLease agreement, amendments and prior schedule
02 · Reasoning2 modes
RO-04

Bright-line threshold applied

A classification test is applied as a hard cutoff instead of the documented judgement required.

RO-06

Remeasurement read as modification

A CPI-index reset is re-priced as a new contract instead of retaining the original rate.

Stage proposesClassification, rate input and confidence
03 · Tool / write2 modes
RO-02

Low-confidence schedule posted

A schedule reaches the register despite insufficient certainty in the extracted terms.

RO-05

CIP depreciated pre-service

Depreciation starts from a project-completion date, not the placed-in-service date.

Stage writesOnly where write access and approval policy allow it
04 · Output1 mode
RO-01

Embedded lease unflagged

A dedicated-fleet or colocation contract reads as a service agreement, and the identified-asset test never runs.

Stage returnsThe schedule the accountant reviews and posts
05 · Change / Version1 mode
RO-07

Silent rate-hierarchy regression

A configuration update applies the risk-free-rate expedient across every class instead of by elected class.

Stage tracksModel, prompt, classification and rate config
Sev-1 · posts outside the boundary Sev-2 · a wrong entry reaches the register Sev-3 · a source degrades, schedule goes to review

Affected slices

One score can hide where the rework concentrates

An overall extraction-accuracy score can look strong while errors concentrate in one asset class. Nestack reports the schedule-rework rate by slice, not only in total.

Slice performance — reported separately, not only in aggregateIllustrative example
SliceFailure rateLift Lift vs. thresholdStatus
Service contracts with embedded leases7.6%4.0× Review
Leases with renewal and purchase options5.0%2.6× Review
Variable-payment and index-linked leases3.2%1.7× Watch
Fixed-payment real-estate leases1.9%0.8× Normal
Bar: schedule-rework rate lift vs. the fixed-payment real-estate baseline · scale 0–4.0× · tick at 2.0× 2 of 4 slices over threshold

Evidence-linked improvement

The loop closes when the schedule is retested, not explained

A cycle ends when the wrong remeasurement is a case the next release has to pass. That suite is what the next schedule built is measured against.

Improvement cycle · five stagesSwitchback — the path turns at Improve and returns at Learn
01Detect

Schedule-rework rate rises in an asset-class slice.

02Diagnose

The schedule that quietly reset traces back through its clauses until the cause narrows to one.

03Improve

Version-stamp the correction and attach the schedules that exposed it.

04Verify

The affected cases run again, and a fail stops the release.

05Learn

It becomes a permanent test, and the classification rules move with it.

Learn → DetectThe return edge. The next schedule built is checked 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, schedule-drafting workflow, evaluation, integration, then production validation and handover.

Workstream Week 1Week 2Week 3Week 4Week 5Week 6
01Lease-workflow discovery and the automation boundary.
02Lease-system and document-source review.
03Classification-test and rate-hierarchy policy mapping.
04Lease-term ingestion and normalisation.
05Drafting logic and lease detection.
06Confidence scoring and flag routing.
07Controller approval workflow.
08Register and ERP integration.
09Remeasurement and classification cases.
10Guardrails and posting controls.
11Schedule-trail instrumentation.
12Deployment, documentation and Agent Care handover.
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 tier PilotOne lease, one office ProductionProduction register integration AdvancedMultiple entities / portfolios
Introduced at Pilot
Drafting to your terms and rules
Controller approval
Extraction-accuracy baseline
Introduced at Production
Reporting by asset class
Approval workflow in your systems
Approved write-back
Asset-register integration
Introduced at Advanced
Multi-standard lease rules
Multi-stage controller approvals
High lease volume
Multi-standard lease controls
Build price From $5,000 From $8,000 Custom quote
Final build priceConfirmed after discovery based on integrations, workflow complexity, lease volume, approval controls and deployment requirements.
Separate from buildBuild 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.

You bringWe build with it
01Your lease agreements and prior schedules Lease-record ingestion and clause mappingWeek 1
02Representative executed leases and amendments Term ingestion and normalisation, and the schedule baselineWeek 2
03Your classification policy and rate-hierarchy elections Classification-test and rate-hierarchy mappingWeek 1
04Access to relevant APIs, feeds or exports Lease-system and document-source assessment, then integration setupWeek 2
05Schedules you would not want posted Embedded-lease cases and the evaluation suiteWeek 4
06What no schedule may infer Confidence scoring, flag routing, guardrails and posting controlsWeek 3
07Named lease accountants to review schedules Controller approval workflow, then pilot and production validationWeeks 5–6
Nothing else is required Deployment, documentation and Agent Care handover are ours.

Delivery timeline

Four phases across six weeks

The bands follow the work as it actually falls, so week 5 holds evaluation and pilot at once.

Phase W1W2W3W4W5W6
Discovery W1
Build W2 – W3
Evaluate W4 – W5
Pilot & Launch W5 – W6
Week focus W1Lease workflow discovery, classification mapping and the automation boundary W2Source integration and the schedule-drafting baseline W3Drafting workflow, confidence logic and approval controls W4Classification, remeasurement and impairment case testing W5Register integration, pilot leases and targeted corrections W6One close cycle posted under the financial controller, then Agent Care handover
Reading the bandA band covers the weeks its work is named in, and no others. The week 5 overlap is real, not padding.
At the end of W6Once the cycle validates, monitoring moves to Agent Care.
DurationSix-week plan shown · typical delivery 4–6 weeks depending on scope confirmed in discovery.

Next step · Accounting AI agent

Build a lease and fixed-asset agent around your controller's approval chain.

Show us your lease agreements, your classification policy and who signs off. You get terms extracted, a schedule built against a rate you approve, and a register your controller can post from.

Nestack Agents · Lease & fixed-asset accountingAGT-ACC-19 · Agent Care available after launch