The banking API that decided the most radical thing it could do was make money movement boring - and then quietly moved hundreds of billions of dollars.
Here is a thing about moving money in the United States: it mostly works. Payroll clears. Cards charge. The Automated Clearing House shuffles trillions of dollars around in files that are, and I cannot stress this enough, CSV files. The system is a marvel of mid-century engineering that has been quietly load-bearing for the entire economy for decades. And then, right at the last mile, it is missing a little something.
That little something is the whole business of Increase. The company, founded in 2020 by Darragh Buckley, sells "enterprise-grade banking APIs" - a single programmatic interface to ACH, wires, real-time payments, FedNow, checks, and cards. The pitch is not "we will reinvent banking." The pitch is closer to: the rails are fine, but the part where the data reaches you is broken, and we fixed that part.
Buckley's résumé is the kind that makes other engineers nod. He was, per his own account, the first employee at Stripe, joining in 2010. He then spent years watching the same frustration recur - that the interesting, useful data coming out of the underlying financial networks kept getting flattened before anyone downstream could use it. "It's hard to build good, reliable product on top of an email failure flow," he has said, which is both a technical complaint and, if you have ever worked in payments operations, a small cry for help.
So Increase does an unglamorous thing very deliberately. It passes the underlying network data through to its customers, more or less untouched - unique trace identifiers, real settlement information, the works. It built its own bank core rather than renting one, specifically to control the constraints. And it treats reliability not as table stakes but as the product itself. "We just try to do all of our little pieces, well, reliably, and transparently," Buckley says. It is a sentence engineered to be unquotable, which is presumably why customers like it.
The economics are the genuinely unusual part. Increase never raised a proper venture round. In a category where competitors have absorbed hundreds of millions of dollars of investor money - and where at least one high-profile peer imploded spectacularly - Increase is self-funded and, as of 2025, profitable. Over the prior year the company says its monthly payment volume tripled and deposits grew fourfold. Its customers are a fintech honor roll: Ramp, Check, Pipe, and others quietly running their money movement on top of it.
- Filed from the Fintech Desk. Facts drawn from public sources; figures are company-reported and approximate.
Increase exposes the core U.S. payment rails as clean API primitives. A company can open bank accounts, send and receive money across every major network, issue cards and checks, and reconcile all of it - without wiring up direct connections to a bank itself.
Origination and receipt with direct network access and unique trace identifiers for clean reconciliation.
Send and receive domestic Fedwire transfers programmatically.
Always-on payments over the RTP network - now 70%+ of U.S. distribution.
API access to the Federal Reserve's instant payment service.
Virtual and physical card issuance, plus push-to-card payouts.
Physical check issuance and deposit, handled through the API.
Programmatic accounts with custom account numbers for fund management.
Transaction tracking and transparent pass-through of underlying network data.
The tell of infrastructure done right is that you rarely see it. These companies move money on Increase so their own users never have to think about the plumbing.
The founder. Darragh Buckley, an MIT-trained engineer, was Stripe's first employee before leaving to build Increase. His philosophy is almost aggressively unfashionable: finish your own work before chasing shiny things, and treat "boring, unexciting, reliable" as the highest compliment a payments product can earn.
The odd sideline. In 2025, Buckley disclosed a stake of more than 10% in Twin City Bank of Longview, Washington - his third investment in a Washington community bank. His thesis: community banks' real strength is "their relationships and knowledge," an asset the market undervalues. Reportedly, a rival even hired a PR agency to seed negative stories about the deal, which is its own kind of compliment.
The moat. No outside capital, a self-built bank core, and a customer base that stays because the money keeps arriving on time. In fintech, that last part is rarer than it should be.
TechCrunch reports Buckley acquired a >10% stake in Twin City Bank, receiving FDIC "non-objection for control." His third Washington community-bank investment.
In a Soar Payments interview, Buckley details the reliability-first philosophy and the decision to build Increase's own bank core.
Increase reports profitability - monthly payment volume tripled, deposits up fourfold over the prior year.
Increase founded and joins Y Combinator's Summer 2020 batch.
DOSSIER: INCREASE · FINTECH DESK · SOURCES: INCREASE.COM, Y COMBINATOR, TECHCRUNCH, CRUNCHBASE, SOAR PAYMENTS