Two companies set out to replace the same piece of software this week, and investors priced the bets in opposite directions. Maximum, a US start-up building what it calls an AI-native operating system for banks, came out of stealth on August 3, 2026 with a $30 million seed round led by CRV. A day later 10x Banking took £40 million from AshGrove Capital in a mix of equity and debt after its first profitable year. Same target — the core banking system of record — but one raise is a cheque against a thesis and the other is a cheque against a book of business.
That target is the graveyard of enterprise fintech, and the arithmetic behind it is softer than the pitch decks suggest. Maximum’s launch materials put the prize at “nearly five thousand” US banks, more than 70% of which run cores built in the last century. The FDIC’s first-quarter 2026 Quarterly Banking Profile counts 4,278 insured commercial banks and savings institutions, and that figure falls every quarter through consolidation. The market for a rip-and-replace core is real, but it is shrinking while the number of vendors chasing it grows.
What Maximum actually raised, and on what
Maximum was founded by Randy Fernando, who sold Vault to Acorns in 2017 and Power to Marqeta in 2023. CRV led the seed with Pear VC, Restive, Plug and Play Ventures and Anthemis participating. The product is a real-time ledger with agents a bank can build and deploy against its own operational workflows, plus controls aimed at AI-era fraud. Fernando told Forbes the company will do full migrations only — no sidecar deployment running alongside an incumbent core — and will finish one bank before starting the next. There is no public product and no named bank customer yet.
That is a deliberately slow route into an unusually concentrated market. Fiserv serves roughly 42% of US banks, Jack Henry 21% and FIS 9%, more than 72% between three vendors, with around 20 firms splitting the remainder. A conversion typically runs 18 to 24 months, consumes internal staff a community bank does not have spare, and triggers exit penalties on the way out. Anyone who has scoped one knows the general ledger is the last thing a bank moves, not the first.
The incumbents have receipts; the challenger has a thesis
No incumbent has responded to Maximum, because there is not yet anything to respond to. The more instructive response is 10x Banking’s, which arrived with numbers: EBITDA-positive, more than 10 million live customer accounts, over 10 new financial institutions onboarded and annual recurring revenue up more than 30% across the past 12 months, serving Westpac, Chase UK, Old Mutual and West Brom Building Society. Growth debt is available to that profile. It is not available to a pre-product seed, which is precisely why the two raises are structured differently.
“For decades, banks have attempted to serve the evolving needs of customers on infrastructure that was never designed for today’s world,” Fernando said. Caitlin Bolnick Rellas, General Partner at CRV, framed it as a generational reset: “Every generation of infrastructure eventually reaches a point where incremental improvements are no longer enough.” Antony Jenkins, Founder, Chair and Chief Executive at 10x Banking and a former Barclays chief executive, makes a narrower claim in 10x’s own announcement: institutions “have a clear ambition to innovate, but many remain constrained by infrastructure that was not built for real-time, digital banking.” Constrained is not the same word as replaced.
What the regulator says about an LLM near the ledger
Almost nothing, which is the problem. On April 17, 2026 the Federal Reserve, FDIC and OCC issued revised interagency model risk management guidance — SR 26-2, published by the OCC as Bulletin 2026-13 — superseding the 2011 sound practices that governed the field for 15 years. It carves out the exact technology being sold: “Generative AI and agentic AI models are novel and rapidly evolving. As such, they are not within the scope of this guidance.” A request for information on banks’ AI use is promised but unpublished. The guidance is also “expected to be most relevant to banking organizations with over $30 billion in total assets” — which describes almost none of the community banks a challenger core realistically wins first.
So a $2 billion community bank deploying agents against its own money movement has no current supervisory framework to cite at an exam and no peer to point to. That is not permanent, but it is why the first Maximum migration will take longer than the engineering does — the same adoption drag Ambrook found selling software to first-time digitisers, transposed into a regulated ledger. Expect capital to keep splitting the way it did this week: growth debt to whoever can show a migrated book, seed equity to whoever argues the ledger itself has to change. Both can be correct, but only one has to survive a conversion to prove it — and core banking has buried better-funded attempts than the neobank cohort that learned licensing is the easy part.
Rick Steves
Rick Steves has seen business and economics through many lenses. He joined the financial services industry in 2009, and has been a financial journalist since 2011. He holds a degree in Business Administration and has experience producing real-time news, from both buy-side and sell-side, as well as for retail traders, brokers and service providers. Steves' work has appeared in a variety of online publications including FX Street, NewsBTC, FinanceFeeds, and The Industry Spread. Rick has great interest in the dynamics of the trading industry. The never-ending clash between technology, economics, regulation, and more importantly, the people.
