Charter Hall: Reducing Manual Invoice Work Across a Property Portfolio

An invoice can look simple on the page and still take time to approve. Someone may need to find the matching purchase order, check the supplier and the property, compare amounts with agreed terms, and resolve a mismatch before payment.

Charter Hall, Australia's leading diversified property funds manager, is using AI for invoice processing and contract validation, two finance workflows that draw on records from property operations. Together, they make a focused story about applying AI to recurring checks, then measuring the work those checks can return to the team.

Why invoice processing needs operational context

An invoice rarely contains all the information needed to judge a charge. The reviewer may need to connect it to a purchase order, a quote or a rate card, then check the supplier, the entity, the dates and the amounts. If one invoice covers several orders or properties, the team may also need to allocate its lines before it can move forward.

At portfolio scale, these checks repeat across many invoices. A person can spend time reading documents, searching for records, entering data and following up on exceptions. The workload grows with volume, while finance teams still need a clear reason for each decision. The small errors that slip through at this scale are where margin leakage in FM invoices accumulates, invisible on any single document and material across a year.

Before the platform, Charter Hall's portfolio ran largely on asset-level spreadsheets, scattered files and manual follow-up. The CMMS, won on competitive tender, became one operating layer across the portfolio and the record the finance agents now work on.

Use case 1: Automate the first pass on invoices

Facilio's Invoice Processing Agent is designed to handle the repeatable steps between invoice arrival and a finance decision. It reads submitted documents, extracts key fields, and checks them against configured operational and commercial records. Because purchase requests, vendor quotes and invoices already sit together in Facilio's procurement and billing automation, the agent checks an invoice against the order it was raised from rather than against a separate system.

A typical workflow moves through four stages:

This workflow reduces repeated data entry and record searching on invoices that pass the configured checks. It brings exceptions to a reviewer sooner and gives vendors a clearer explanation when something needs correction. The team can spend more of its time on mismatches and decisions that need judgment. The checks themselves are the same ones that make AI invoice validation reliable in any facilities operation: a named record to compare against, and a defined outcome when the values differ.

Use case 2: Bring contract terms into the review

Contract validation is also live at Charter Hall. That broadens the story from invoice handling to the commercial checks that support property operations.

When a team configures contract validation to compare submitted work or charges with agreed terms, it reduces manual lookup and surfaces mismatches earlier. The value depends on which contracts, fields and rules are in scope, and on how the workflow sends exceptions to a reviewer. The comparison is only possible when rates, scopes and service levels are kept current on the vendor record rather than in a filing cabinet.

For another property team, start by identifying where contract information lives and which checks follow stable rules. Choose a small, high-volume contract set. Agree on what counts as a match, which variations need human review, and what evidence the reviewer should see.

What value can the workflow create?

The product flow points to operational benefits a team can test:

• Less repeated data entry and less time spent locating the same records.
• A faster first review for invoices that pass the configured checks.
• Earlier visibility into mismatches and exceptions.
• Clearer reasons for vendors and internal reviewers when an invoice needs correction, which is what shortens the rework cycle behind most FM invoice disputes.
• More staff capacity for decisions, supplier coordination and financial oversight.

These are the process effects the workflow is designed to produce. A customer story should show which effects occurred and how the team measured them.

Estimate the time a workflow could return

A simple estimate starts with the work a person performs today and the time still needed after automation:

Potential hours returned per month = eligible invoices per month × (manual minutes per invoice less assisted minutes per invoice) ÷ 60

For example, if a timed review step falls from 50 minutes to five, that step uses 45 fewer minutes per invoice, a 90% reduction in handling time for that step. Multiply the difference by the number of invoices eligible for the workflow to estimate the gross time involved.

Keep people in control of exceptions

Automation needs clear boundaries. The Invoice Processing Agent provides configurable checks, a path for soft flags that need human review, rejection handling for hard failures, and an audit trail for decisions. These controls let teams see what the agent checked and why it routed an invoice for review.

A credible customer story explains which decisions the agent handles, which remain with finance or property teams, and how someone can inspect the record.

How another portfolio can apply the same approach

A practical rollout begins with the workflow and its records. Before automating, map what arrives, where the data needed for validation lives, and who owns each exception.

• Choose one recurring workflow with enough volume to matter and rules that can be stated clearly.
• Map the source documents and records the agent needs to compare.
• Agree on pass, review and rejection conditions with finance and property stakeholders.
• Measure the current process before deployment, including handling time, exception volume and rework.
• Test with representative examples, review false flags and missed issues, then tune the rules before expanding.
• Track the same measures after launch and report results against the original baseline.

This gives a team a way to evaluate value before expanding. It also makes the result easier to explain: Which work moved to the agent, which exceptions stayed with people, and how the workflow changed the team's workload. Facilio's FM Ops Engineers run this process with the customer team on the same agentic platform that hosts the ready agents, and they stay through rollout.

Frequently asked questions

Talk to us about your portfolio

Every invoice checked against the record it was raised from, with people reviewing only the exceptions. That is the shape of the work at Charter Hall, and the shape it can take on another portfolio.