Accounts payable is the function most small and medium businesses automate last and complain about first. It is repetitive, it is high volume, it carries genuine fraud exposure, and the tools available to smaller organisations have generally handled it by scanning a document and extracting a total.
Process AI Pty Ltd, founded by Brisbane Chartered Accountant Benjamin Whitehouse, has taken a different position on what accounts payable automation should do. The company has developed and commercially launched an AP automation platform for Xero, built around the premise that the useful work happens below the invoice header.
Line item processing rather than header extraction
Conventional invoice capture tools read the fields that are easiest to find: supplier name, invoice number, date, total amount. That is sufficient to get a bill into a ledger. It is not sufficient to know what a business actually bought.
The Process AI platform processes every line item on an invoice. The difference is consequential for anyone who needs to reconcile purchases against orders, allocate costs across projects or cost centres, or identify variances between what was ordered and what was billed. A header total confirms an amount is owed. Line item data confirms whether the amount owed is correct.
For an accountant reviewing a client’s records, the distinction determines whether the ledger supports analysis or merely records outcomes. Benjamin Whitehouse’s design decision here traces directly to advisory experience: the questions a business needs answered about its expenditure cannot be answered from summary data.
Purchase order management inside the workflow
The platform also manages purchase orders, closing the loop between commitment and payment. Where a purchase order exists, invoice line items can be matched against it, and discrepancies surface before payment rather than during a subsequent review.
Purchase order discipline is common in larger organisations and uncommon in smaller ones, generally because the administrative cost of maintaining it has outweighed the benefit at lower transaction volumes. Automating the matching step alters that calculation. The control becomes available to businesses that could not previously justify the overhead of operating it manually.
Supplier identity and bank account matching
The platform applies intelligent supplier identity and bank account matching, and this addresses a category of risk that has grown considerably for Australian businesses.
Payment redirection fraud, commonly executed through phishing, works by presenting a business with a plausible request to update a supplier’s bank details. The invoice looks correct. The supplier name is correct. The amount is correct. Only the account number has changed, and the change is discovered when the legitimate supplier follows up on a payment that never arrived.
Manual accounts payable processes catch this inconsistently, because the person entering the payment has no reliable reference for what the supplier’s details were previously and no systematic prompt to check. Automated identity and account matching applies that check on every transaction rather than on the transactions someone happens to scrutinise. The same mechanism reduces ordinary data entry errors, which are more frequent than fraud and cumulatively expensive to correct.
Why the Xero ecosystem
Xero is used widely across Australian small and medium businesses and the accounting practices that serve them. Building for that ecosystem rather than for enterprise resource planning platforms reflects a deliberate targeting decision by Benjamin Whitehouse: the automation gap is widest not among large organisations, which have had access to sophisticated AP tooling for years, but among the businesses beneath that threshold.
An SME processing several hundred invoices monthly has the same categories of problem as an organisation processing tens of thousands. It has less capacity to absorb the cost of getting them wrong.
What is being built next
Benjamin Whitehouse is currently developing a fully autonomous AI accounting and analytical system, designed without a traditional user interface, for small and medium enterprises and insolvency professionals.
The absence of an interface is the substantive claim. Accounting software has generally been built on the assumption that a person operates it: opening screens, entering data, running reports, interpreting output. An autonomous system inverts that relationship, performing accounting and analytical work continuously and surfacing conclusions rather than waiting for a user to request them.
For insolvency professionals in particular, the application is specific. Assessing the position of a business under financial pressure requires rapid reconstruction of financial reality from records that are frequently incomplete. Analytical work that currently consumes days of professional time is a candidate for automation, and the value of compressing that timeline is highest precisely when a business has the least time available.
The advisory basis for the product
Process AI is the work of a Chartered Accountant with 32 years in the profession rather than of a software company selecting accounting as a market. Benjamin Whitehouse founded the Viden Group, where he serves as Founder, CEO, and Director across advisory, investment, and development entities, and the product decisions at Process AI carry the marks of that background: line item detail because analysis requires it, purchase order matching because control failures are observable in client files, and fraud checking because payment redirection has become a live risk for Australian SMEs.
Details on the Viden Group and its activities are available at viden.
