GrailPay: Building the Trust Layer for B2B Payments

All human wisdom on the ACH rail has been contained in two words: wait and hope. GrailPay offers a third: know.

Every year, more than $93 trillion moves across the Automated Clearing House (ACH) network, roughly nine times what Visa and Mastercard process combined. Yet despite the scale and importance of ACH, businesses still have limited ability to know whether a payment will successfully clear before it is initiated.

That’s been true for the last 50 years. Mistyped account numbers, a payment sent to a closed account, insufficient funds, or outright fraud can all cause an ACH payment to fail, and the business on either end of the transaction won’t find out until one to three days later.

The ACH network was built in the 1970s, primarily to move payroll and bill payments between banks in batches, and the rail has changed very little since then. There are workarounds to gain some assurance that a payment will succeed, but all of them are clunky, add friction for payer and payee, and still don’t tell you whether a specific payment is going to be successful:

• Micro deposits — Send a couple of test cents to an account, then wait a day or two for someone to confirm the amounts before trusting it.
• Prenote — A zero-dollar test transaction that confirms an account number is formatted correctly and catches fat-finger errors.
• Database lookup — Services that offer a static, backward-looking check to tell you whether an account in their data set has been reported closed or fraudulent.

Now compare this to the card networks. Cards provide real-time authorization at the point of purchase. ACH offers no equivalent, just a return window that has to pass before you know a payment actually cleared. Despite that gap, ACH isn’t going anywhere and is still the most cost-effective way to move money. That holds especially true for business-to-business (B2B) payments, where volume is outpacing every other segment, growing nearly 10 percent year-over-year.

That bend-but-don’t-break tolerance has held up for decades. But as payments settle faster and humans get taken out of the loop entirely, relying on infrastructure built in the 1970s is quickly becoming untenable:

• Adoption of instant payments — FedNow and RTP settle in seconds, not the one-to-three-day window a standard ACH return relies on, and once the money moves, it can’t be pulled back. Risk decisions that used to happen after the money moved now have to happen before the transaction.
• Proliferating fraud — AI is enabling fraudsters to spin up synthetic identities, run business email compromise scams, and take over accounts outright, moving money out within hours, long before an old-style return code would ever catch up.
• Rise of agentic commerce — By 2028, Gartner predicts AI agents will handle 90 percent of all B2B purchases, moving more than $15 trillion through machine-to-machine exchanges. As agentic commerce develops, a growing share of transactions will increasingly be initiated and executed by software rather than humans.

Enter GrailPay.

GrailPay is building the risk intelligence layer that ACH never had. Rather than confirming an account exists, or waiting for a return code to find out what already happened, GrailPay predicts whether a specific payment will actually clear before the money moves.

Our conviction in GrailPay grew as we spent time with Will Messina and his team, studied the product and market, and evaluated the limitations of existing solutions. That conviction led MissionOG to lead the company’s $10.5 million Series A, and it centered on a few key elements of our investment thesis:

• Proprietary data flywheel. Most account validation tools leverage the same third-party bank data. GrailPay built its own ACH processing operation first, handling billions of dollars in annualized volume and creating a proprietary dataset of payment outcomes. Every payment that clears, returns, or gets flagged feeds back into the models. The more transactions GrailPay sees, the smarter its models get.
• Prediction, not just validation. Legacy tools answer whether an account exists and is in good standing. GrailPay answers whether this specific payment will actually go through, a different and much harder question, and increasingly the one that actually matters.
• Capitalizing on secular shifts in payments. Real-time rails, new NACHA fraud rules, and the early stages of agentic commerce are all converging on the same requirement at once: payment risk must be evaluated before a payment is initiated rather than after. GrailPay’s real-time risk engine, trained on outcome data from its own processing volume, is built to evaluate that risk before funds move.
• Team domain depth. Will Messina built GrailPay’s platform around a critical insight from dozens of conversations with B2B businesses who kept seeing their ACH payments fail but couldn’t explain why. He’s complemented by his cofounder, Lee Jones, whose 30-plus years of enterprise engineering experience built the differentiated platform architecture that makes GrailPay’s data model possible. Together, they bring the rare combination of payment and risk domain expertise, the ability to execute on solving this unique problem, and a vision for what the future of payments requires.
• The vision goes well beyond one check. Validating an account is just the start of what GrailPay is building. The vision is a full risk and identity layer for B2B payments: transaction-level scoring on every individual payment, and a portable identity that carries an entity’s trust profile across platforms and rails. That’s where we expect much of the company’s next chapter to be written.

For MissionOG, investing in GrailPay means backing a company that is rebuilding the risk infrastructure underneath one of the largest and most overlooked payment rails in the country. A few months into the partnership, we remain convinced this is one of the more important infrastructure bets in payments today.