Featuring Allison Stites, Paystand
B2B payments involve more than simply sending money from one business to another. Behind every transaction are finance teams managing invoices, payment information, accounting systems, approvals, collections and international transfers.
As these processes become increasingly complex, businesses need to rethink how their finance operations work.
At the NextGen Payments & Cross-Border Innovation event at Techsommet , presented in partnership with Paystand, Allison Stites, Head of Product Marketing and Commercialization at Paystand, explored how two technologies — AI agents and programmable money — are changing what businesses can do.
Her central argument was that these technologies are not simply adding new tools to existing processes. They are creating opportunities to change the way businesses operate.
“For the first time, the work of moving money and the work of running finance can both happen at software speed”
AI Agents: Creating More Capacity in Finance Teams
Allison began by comparing organisational change with her experience practising Iyengar yoga.
In yoga, forcing a movement does not necessarily help someone achieve a pose. Progress comes from changing the process and creating more room for movement. She connected this idea to business operations: instead of forcing new technology into existing processes, businesses can use it to rethink how work gets done.
At Paystand, AI agents are referred to as digital employees.
Allison described how AI has evolved over the past few years. Businesses moved from chat-based tools, where users entered prompts and received responses, to copilots that could help complete workflows and build code. The next development is AI agents with defined job descriptions, functions and boundaries, allowing them to perform tasks autonomously.
For finance teams, this creates an opportunity to reduce repetitive work.
Consider the everyday process of understanding outstanding customer payments. A finance employee may need to export information from an ERP, check emails to find the last customer communication, review bank statements and compare information across different systems.
This work can be time-consuming, especially when payment descriptions do not clearly identify the related invoices.
AI agents can take some of this work off the team’s plate, reducing the need to manually collect and compare information.
Programmable Money: Moving Beyond Traditional Payment Processes
Allison explained that AI agents are only one part of the changing B2B payments landscape. The other is digital and programmable money.
She discussed stablecoins and how programmable money can change the way businesses move and manage funds.
One important difference is settlement speed. Rather than waiting business days, final settlement can happen in seconds. Cross-border payments can also move much faster, with the ability to program when money is released.
This changes the focus from simply processing transactions to governing how money moves.
However, faster payments alone do not solve every problem.
Allison pointed to the fragmented technology stacks used by finance departments. Businesses may have different vendors and tools for accounts receivable, accounts payable, expense management, treasury and reporting, all connected to their ERP systems.
Even sophisticated systems can require significant manual work. Someone still has to fix broken connections, update labels, find information in emails and investigate whether a customer has paid.
Adding AI to these disconnected systems may speed up individual tasks, but it does not automatically solve the underlying problem.
Connecting Payments With Financial Information
Allison presented programmable payment rails as a way to address this disconnect.
When businesses move from legacy payment rails to programmable ones, payment information can remain connected to its accounting context. This gives AI agents more useful information to work with.
She explained that traditional finance processes can involve multiple vendors, transaction costs and waiting periods before cash reaches the business.
With programmable money and AI, businesses can work towards more connected payment processes, faster movement of funds and more continuous financial reporting.
The goal is not simply to automate individual tasks. It is to reduce the manual investigation and data entry that finance teams regularly handle.
How Paystand Is Applying These Technologies
During the presentation, Allison introduced Paystand as a business payment network. It focuses on helping B2B payments by eliminating transaction fees, automating manual financial processes, and accelerating settlement and cash velocity.
She explained that Paystand’s solutions cover accounts receivable, accounts payable, expense management and global payouts, with digital employees being introduced across these areas.
On the accounts receivable side, Allison described a reporting agent that can query receivables and payment data. It can present an at-risk ageing accounts list and provide real-time cash forecasts.
She also discussed a collections agent, then in private beta, that can draft communications and send them to customers as part of the follow-up process.
For spend management, she introduced an agent that integrates with Slack and Teams. Employees can use it to make purchase requests and submit expense reports, while controllers define the approval policies in advance. The agent can also collect receipts, code expenses and route reports for approval.
The presentation also covered borderless payouts. Businesses can pay international vendors and global workers in their own currencies, while reducing the need to send individual international wire transfers and investigate unclear foreign exchange costs or delayed payments.
The Bigger Opportunity: More Strategic Finance Teams
Allison concluded by returning to the idea of increasing the range of motion within finance teams.
As repetitive work such as data entry, investigation and comparing information across systems becomes simpler, finance professionals can spend less time managing manual processes.
Her point was not just that technology can make existing work faster. It can create capacity for businesses to take on more ambitious problems.
The presentation highlighted a connected direction for B2B payments: AI agents handling defined finance tasks, programmable money supporting faster and more connected transactions, and businesses rethinking the processes that sit behind their payments.
For Paystand, these technologies are part of a broader approach to making business payments and finance operations more connected.
Q&A With Allison Stites — Original Event Conversation
Can AI agents make material B2B payments without human approval?
Moderator: “How close are we to AI agents actually having authority to make material B2B payments without human approval and what controls do you think will be non-negotiable before that happens?”
Allison Stites: “I’d ask like how one defines material. But I think the, for example our spend agent is already making decisions over virtual card purchases and physical card purchases. You know, you can set an approval rule. You can set a rule that’s like if it’s a below $5,000 auto approve.
In the instance of like releasing a million-dollar payment, I would ask the business like, is that something you want a digital employee to do? If that is like, would you have an AR analyst do a million-dollar payment?
I don’t think that the rise of digital employees means that you should think about them as actually staff. And so if you’re going to bring on staff, what degrees of freedom do you give them?
And I wouldn’t treat a digital employee as a senior CFO. Like we all know they’re not like a corpus, they don’t have a corpus body. They don’t have any context.
So I think that’s a decision for businesses to decide and to make based on their own business policies.
And for the vendors of B2B payment solutions that are integrating this technology, I think you should definitely be discerning and ensure that they give you the option to set the policy. They don’t set it for yourself.
At Paystand, we would never release a digital employee that’s going to execute payments without your explicit definition of what they can do.”
Where does the real opportunity in programmable payments lie?
Moderator: “If the payment becomes programmable, does the real opportunity lie in making the payment itself cheaper or in embedding payment logic directly into commercial contracts, invoices, and supply chains?”
Allison Stites: “Both. Both. But the second, like when you can make the payment programmable and you can configure logic based on invoice timing.
A lot of the cost of payments today is around the message, the communication layer between banks and credit card networks. That’s what you’re paying for. It’s the risk and the compliance.
And if you have that baked into invoices where there’s been counterparty risk assessed, you’ve validated the recipient and the sender, that’s what removes a lot of the cost.”
What do audit trails and SOC controls look like with autonomous agents?
Moderator: “If AI agents are effectively owning accounts receivable and collections end to end, what does the audit trail look like? And how are teams satisfying SOC controls when an autonomous agent is making settlement decisions rather than a person?”
Allison Stites: “So, I actually think audit trails become much more comprehensive and clear because an AI agent can record instantaneously actions that happen.
There’s no remembering to document. There’s no things lost in an email. It’s recorded in a clean audit trail that’s going to be exportable, associated with the dashboard that you have, at least in Paystand.
So I think you’re going to get a more detailed audit trail with agents than you do with humans, at least for the areas that the agent is responsible for.
And then with the SOC control, similarly, you don’t just spin up. There’s a policy framework that digital agents are allowed to operate within.
So, they’re going to be interrogating whether the agent complied to the policy framework, and if it didn’t, then that’s a bug in the system.”
Watch the Full Event Recording
Want to explore more insights from the NextGen Payments & Cross-Border Innovation event?
Watch the full event recording on YouTube to hear the complete presentations and discussions on AI, B2B payments and cross-border payment innovation.
