Use cases · Real estate

Net operating income, reconciled across the books, the contracts and the meters.

A real-estate question answered across disconnected systems — with the numbers, the method, and the honest caveats. Figures are illustrative (synthetic data).

Use-case · cross-source reasoning

One question. The AI reads the books, the contracts and the invoices — and finds the money you forgot to bill.

Net operating income is down even though the property is fully let. You just ask why — and the AI pulls four disconnected sources, cross-references them, names the leak, and hands your property manager the fix.

"Vårberg Centrum is at 100% occupancy, so why is NOI (driftnetto) down this year?" asked by P. Persson · Controller
  1. 1

    Decompose the NOI drop

    Splits driftnetto by cost line. One line dominates: electricity, +560 tkr year-on-year.

    General ledger
  2. 2

    Read the lease contracts

    Finds 2 of 9 tenants are contractually liable for their own electricity (sub-metered recharge clause).

    Signed leases (PDF)
  3. 3

    Audit the sent invoices

    Checks every invoice to those 2 tenants — the electricity recharge line is missing for 7 months.

    Sent invoices (AR)
  4. 4

    Quantify & assign

    Adds up the un-billed recharges, sizes the NOI hit, and writes a task list to recover it.

    Synthesis

The paradox isn't operations — it's revenue leakage. Costs rose because electricity went up, but two tenants who are contractually billed for that electricity were never invoiced for it. The "missing" NOI is real money sitting un-billed.

NOI (driftnetto) trend

·  the paradox

Quarterly net operating income, Vårberg Centrum. Occupancy held at 100% the whole period — yet NOI fell.

4,2M 4,0M 3,8M 3,6M 4,10M 3,98M 3,92M 3,71M Q1 Q2 Q3 Q4 Occupancy held at 100% −9,5%

What moved the NOI

·  cost bridge

Year-on-year walk from last year's NOI to this year's. Rent and other costs barely moved — electricity is the whole story, and most of it should have been billed back.

4,10M NOI −1 yr +46 Rent +124 Other OPEX −560 Electricity −248 3,71M NOI now
Recoverable now · back-charge 248 000 kr

Electricity the two tenants consumed and owe under their leases, but were never invoiced for across the last 7 months — recoverable in full, and the single largest piece of the NOI gap.

Electricity: paid out vs billed back

·  the gap

For the two recharge-liable tenants only — what the property paid the grid vs what it invoiced the tenants.

0 250 tkr 248 tkr Paid to grid 0 kr Recharged nothing billed gap = 248 tkr

The evidence, side by side

·  contract clause vs invoices sent

Two named leases. The contract says the tenant pays; the invoice history says they were never charged.

Tenant / lease What the contract says What was invoiced Owed
Nordica Café AB
Lease #VC-204 · LOA 240 m²
Tenant pays own electricity by sub-meter, invoiced monthly. Lease §6.3 · p.4 No el-line · 7 mo 142 000 kr
Tandvård Syd HB
Lease #VC-118 · LOA 165 m²
Tenant reimburses actual electricity at cost, billed quarterly. Lease §5.1 · addendum B No el-line · 2 qtrs 106 000 kr

Action list — handed to the property manager

TWAssigned to Terese Widström
Issue back-charge invoices to Nordica Café AB (7 mo) and Tandvård Syd HB (2 qtrs) for un-billed electricity.
High priority recovers 248 000 kr due in 5 days
Fix the recurring billing template so the electricity recharge line auto-applies to both leases going forward.
High priority stops the leak repeating
Verify sub-meter readings and allocation keys for both units against the grid invoices before re-billing.
Medium accuracy check
Set a monitoring alert: flag any month where recharge-liable tenants are billed energy but no electricity line is sent.
Medium reporting → monitoring
From a single question to four assigned tasks — the moment reporting becomes action. Push to your task system

Honest by design: figures are sample data — realistic and internally consistent, not a real property. Every number above is traced to its source: the general ledger (driftnetto by account), the signed lease PDFs (recharge clauses, cited by section), the sent-invoice / AR records (the missing electricity lines), and sub-meter readings — all read in your own tenant, permission-aware and logged per user. The AI quantifies and proposes; a person reviews and books.

See this on your data Browse the use-case library Rung 3 of the capability ladder — document intelligence, cross-referenced.
Use-case · Real estate · Transaction due diligence

A complete DD package in hours, not days.

When a buyer sends a due-diligence checklist, someone has to hunt every document across four disconnected systems before the sale can move forward. AccuraSee Gateway interprets the checklist, retrieves each underlying document from wherever it lives, cross-checks figures across sources, and surfaces every mismatch and gap — so a person reviews a complete, cited package instead of assembling one.

Data sources Lease & rent contracts Rent roll & financial records Inspections, certificates & supplier invoices
AccuraSee Gatewayinterpret · fetch · cross-check · cite
Output A complete, cross-checked DD package + flagged exceptions + a missing-docs list
Document-assembly time Days hours A complete, cited package ready for human review.
Every figure cited
linked to the exact source document & section
A person decides
a reviewer approves the full package before it reaches the buyer

The package, item by item

Buyer's DD checklist · synthetic data
DD package · mid-size transaction
7 objects · assembled from 4 systems
6 of 6 located3 clear · 3 flagged
Lease agreements — all current tenants
Lease & contract system
Found — 12 leases retrieved
Contracted rent vs. rent roll reconciliation
Lease system + rent roll
Mismatch on Object 4 — 1,450 vs 1,395 kr/m²
Energy performance certificates
Compliance document store
Missing for 2 of 7 objects
Option clauses and expiry dates
Lease & contract system
One option expired March 2026 — not renewed
Service & maintenance contracts (active)
Supplier invoice archive
Found — 8 contracts, all current
Environmental inspection reports (5 yr)
Inspection report archive
Found — latest dated Nov 2024
Package assembled — ready for human review 3 exceptions to clear first

Honest by design: AccuraSee Gateway assembles and flags — it doesn't decide. A person reviews the complete package and approves it before it reaches the buyer. All figures are illustrative (synthetic data) — a plausible scenario, not a specific transaction or customer.

Use-case · Real estate · Invoices vs. contract

Catch invoice drift before it silently compounds.

Technical-management contracts spell out every rate, scope and cap — but invoice lines rarely get checked against them systematically. AccuraSee Gateway reads both sides and flags every deviation, line by line, before it becomes a habit.

Data sources Supplier contracts (rates, index, scope) Supplier invoices (lines) General ledger (amounts, cost centres)
AccuraSee Gatewaymatch · check rate & index · flag · cite
Output A deviation list per supplier — flagged lines with amounts and contract references
Recoverable this period +72,300 kr Across 3 suppliers — flagged and cited, for a person to reclaim, query or accept.
18%
of invoice lines flagged this period
+35 kr/h
rate drift over the agreed indexed rate

Line by line — contract against invoice

One supplier shown · synthetic data
Contract saysDeviationInvoice says
Hourly rate + index — electrician
640 kr/h + 3.1% = 660 kr/h
+35 kr/h
As invoiced
695 kr/h
Snow clearing — base scope
Included in fixed monthly fee
Double-billed
As invoiced
Billed as callout · 4,200 kr
Material markup cap
Max 8% on materials
+3 pp over cap
As invoiced
11% applied
Callout > 10,000 kr
Written order required first
Off-contract
As invoiced
14,800 kr · no order found
Fixed monthly management fee
28,500 kr/month
Matches
As invoiced
28,500 kr
Flagged this period · 3 suppliers +72,300 kr

Honest by design: the assistant flags deviations and cites the relevant contract clause — it never withholds payment, raises a dispute, or contacts a supplier. Every flagged line goes to a person who decides whether to reclaim, query or accept. Figures are illustrative (synthetic data); no real customer or supplier is identified.

Have a question in real estate your systems can't answer today?

Bring us a real one. We'll scope a low-risk pre-study and show you exactly what's answerable on your data.

Ask your business data anything. Get an answer in seconds.

Start with a discovery call — we'll scope a low-risk pre-study and show you what's answerable on your data.

What to expect: a 30-minute call · we map your data sources · a live answer on your data — no pressure.