Governed AI agents, explained
The five controls every agent runs under, and how governed agents compare with RPA bots and chatbots.
Read the explainerThe accounts payable agent captures vendor invoices from your AP inbox, codes them using your rules and routes each one to the named approver. It also sets up new vendors with tax and insurance documents collected. Finance approves every payment, and the agent never releases funds.
What it does
Reads each invoice as it lands in your AP inbox, in the layouts vendors actually send, and picks out the vendor, amounts, dates and lines.
Matches it against the purchase order, contract and past invoices, then codes it using your rules.
Flags over-threshold amounts, changed bank details and anything that doesn't match before it goes further.
Sends it to the named approver with the match and evidence attached. Nothing moves until that person approves.
Posts the approved invoice to your accounting system for the next payment run and logs every step.
Where finance approves
These rules are written into the design and approved by your finance team before anything is built.
What it connects to
The agent works in the systems you run today rather than a new platform your team has to learn. We confirm the systems, access and approval path in the written design, and access is granted through your own controls, limited to what the design calls for.
What you get
How it's measured
We size the agent on these measures in the Opportunity Scan and design. After go-live, the monthly report tracks the same measures against your baseline.
Time your team spends receiving, keying, coding and chasing approvals today, and how much of it the agent takes on.
How long an invoice takes from received to approved and ready for payment, and how long a new vendor takes to set up.
What slips today: late fees, invoices sitting unapproved and bank detail changes that need a second look.
The fixed-fee build and monthly run compared with hiring, outsourcing or leaving payables as they are.
Agent or AP software
Is an AP agent just old software with new branding? It's a fair question. Invoice capture, matching rules and approval workflows have been around for years, and AP automation products handle them well for many teams. The difference is less about the technology and more about what you buy, where it works and who runs it.
| Question | AP automation software | A governed AP agent |
|---|---|---|
| What you buy | A software product and its workflow | A service built and run around your workflow |
| Where the work happens | In the product, connected to your accounting system | In your accounting system, AP inbox and approval tool |
| Who runs it | Your team, with the vendor's support | We do, every month: monitoring, tuning and updates |
| Fits best when | Your process fits the way the product works | Invoices vary, exceptions are common or the work spans several systems |
| Ownership | Licensed from the vendor | Source code in your Git, running in your accounts |
If a packaged product already fits your process and your team has time to run it, that may be the simpler answer. For how agents compare with RPA bots and chatbots, see governed AI agents, explained.
Questions
Pricing follows the standard model: a fixed-fee build and a monthly run.
No. Finance approves every payment and the agent never releases funds. It prepares, codes and routes invoices, and a person in finance decides what is paid.
The change stops for a person. Finance reviews it, with the request and evidence attached, before anything is updated.
No. The agent works in your ERP or accounting system, your AP inbox and your approval tool. We confirm the systems and access in the written design.
Some of the building blocks are familiar. What's different is that the agent handles varied invoices, works in the systems you already run, follows approval rules written before it's built, and is operated by us every month. You own the code and the accounts it runs on.
We baseline hours returned, cycle time, risk and leakage, and cost against the alternative during design. The monthly report shows those measures, every approval and every exception against that baseline.
A fixed fee to build and a monthly subscription to run, sized in the Opportunity Scan against the agent's business case. Hosting and model costs run in your own accounts with no markup.
No. It takes receiving, keying and chasing approvals off their plate. Your team still owns the rules, approves what matters and handles anything unusual.
Go deeper
The five controls every agent runs under, and how governed agents compare with RPA bots and chatbots.
Read the explainerScan, design, build and run, with a fixed-fee build, a monthly subscription and a report against your baseline.
See the engagement modelWhy an invoice coded in someone's chat session isn't the same as an agent running in production.
See the side by sideManaged Intelligence Opportunity Scan
In 30 minutes, we'll walk through how invoices and new vendors move today, where finance approves, and what an agent would change. No commitment to build.