Nymbus Found a Smarter Way Into Banking: Build the New Bank Beside the Old One

The most honest sales pitch in enterprise software may be: you do not have to trust us with everything yet. That is the door Nymbus found into American banking. The Jacksonville fintech can replace the central ledger and transaction engine inside a bank. But it can also do something easier to approve: place a new, cloud-based core next to the old one and use it to launch a separate digital brand. No demolition on day one. No ceremonial burning of the mainframe. Just a contained bet.

That distinction matters because a bank core is not another dashboard. It holds customer records, posts transactions, calculates balances and connects the machinery around deposits, loans, cards and compliance. The incumbent market is dominated by Fiserv, FIS and Jack Henry, with newer alternatives including Finxact, Mambu, Thought Machine, Temenos and Alkami in adjacent layers. Switching is expensive, operationally intimate and visible to every customer if it goes wrong.

Nymbus, founded in 2015 by Scott Killoh and Alex Lopatine, packages a cloud-native, API-first core with online and mobile banking, account opening, lending, analytics and marketing tools. It then adds the unglamorous bits that make the software usable: back-office processing, call-center support, compliance help, implementation and growth marketing. Its shorthand is “technology, people and processes.” For a community bank with a lean technology staff, the people and processes may be the actual purchase.

The product is a reversible decision

The sidecar core changes the unit of risk. Instead of asking a board to move the entire institution, Nymbus can help it pick a market - young savers, small businesses, union workers, filmmakers - and build a digital proposition for that group. The new brand gets its own onboarding, products, mobile experience and operating layer while the existing bank continues doing what it did yesterday.

PeoplesBank in Massachusetts is the cleanest version of the playbook. It launched ZYNLO, a digital-first division aimed at younger customers, with Nymbus in 2020. ZYNLO offered familiar online accounts plus a charming hook called Zyng: round a debit purchase to the next dollar and match the spare change. The bank could learn the new system without putting its 135-year-old main institution through an immediate conversion.

Then the trial became a commitment. By 2025, Nymbus said ZYNLO had reached $150 million in deposits and more than 7,500 customers. That June, PeoplesBank moved its main institution - more than $4.4 billion in assets - to Nymbus. The conversion finished a day ahead of schedule, and more than 19,000 customers enrolled on the first day. A five-year test drive is not the stuff of a breathless launch tweet. It is how trust works when the software holds money.

The trick was not making core replacement look easy. It was making the first decision smaller.

Niche banking is distribution wearing a cardigan

The old community-bank boundary was geographic: the people around a branch. Nymbus wants “community” to describe an affinity. CineFi, launched in 2026 with First Entertainment Credit Union, is designed for film, television, music, gaming and creator workers whose income can arrive by project instead of paycheck. Pillur, built with Michigan State University Federal Credit Union, focuses on small businesses. BankMD, launched earlier with TransPecos Banks, targeted medical professionals.

This is not merely changing the hero image and debit-card color. A useful vertical bank needs eligibility rules, products, education, underwriting assumptions, service scripts and partnerships that fit the audience. The approach works when the parent institution has a credible right to serve the niche and a distribution channel into it. It fails when “vertical” means buying ads against a large profession and hoping a purple app creates loyalty.

What failed first

A fair profile cannot turn every migration into a teal arrow pointing up. In May 2022, VyStar Credit Union began what members were told would be a 48-hour interruption while it converted online and mobile banking to a platform built with Nymbus. Access problems stretched for weeks. Members queued online by the tens of thousands, checked balances at ATMs and crowded branches. The underlying core functions and funds remained available, but the digital front door had jammed.

VyStar’s chief executive said the system had been taken down deliberately for adjustments rather than collapsing on its own. Federal regulators later fined VyStar and focused their criticism on the credit union’s deficient planning, testing, contingency preparation and consumer response. Nymbus was the named digital-banking partner, but the public record does not support a tidy, single-vendor villain. It supports a harsher and more useful conclusion: modern architecture cannot rescue weak conversion governance.

What changed minds after that episode was not a slogan. Banks needed staged proof, detailed service levels, nonexclusive contracts, explicit deconversion terms and evidence at larger scale. Nymbus’s public contract sheet now stresses no exclusivity requirement, time-and-materials deconversion fees and negotiated liability provisions. PeoplesBank’s 2025 conversion supplied the flagship proof point the company had been working toward. One success does not erase a prior failure. Together, they show the conditions that separate a pilot from production.

The business behind the bank in a box

Nymbus makes money through enterprise subscriptions and managed services. Public pricing is “on demand,” which is corporate software’s genteel way of hiding the menu. The real cost is broader than a license: implementation, integrations, data migration, parallel operations, testing, training, customer communication and contingency capacity all belong in the budget. Nymbus raised $70 million in a 2023 Series D led by Insight Partners, with The Banc Funds Company, Mendon Venture Partners and clients ConnectOne Bank and PeoplesBank joining. Gesa Credit Union made an undisclosed strategic investment in 2024.

Its competitive difference is breadth with accountability. A bank can buy a core elsewhere, assemble digital channels from specialists and keep operations in-house. Nymbus argues that one partner spanning the core, experience, launch crew and growth engine can shorten the chain of excuses. The trade-off is concentration: outsourcing more work to one platform increases the diligence required on resilience, security, service levels and exit mechanics.

Now, the chat window reaches the ledger

In April 2026, Nymbus launched a Model Context Protocol server for its banking platform. The first version exposes 19 tools for jobs such as customer lookup, account management, money movement and debit-card controls. The compelling detail is not that a banker can talk to software. It is that institutions decide which tools are enabled, which roles can use them and where review is required. Token authentication, encrypted connections, masked personal data in logs and audit trails surround the actions.

That restraint is the product. An AI assistant that drafts a friendly answer is useful; one that freezes a debit card or initiates a transfer is an operator. Nymbus is betting that a standardized, permissioned interface can reduce bespoke integration work while keeping employees in control. It will not work where institutions lack clean data, defined permissions or the patience to start with narrow workflows. “Agentic banking” is a terrible place to discover that nobody owns the approval matrix.

Nymbus fits between the giant core processors and a growing class of cloud-native challengers. It is not a consumer neobank, although consumers touch brands running on it. It is not simply banking-as-a-service, because it can power a chartered institution’s main core. The cleanest description is a full-stack banking platform with a managed launch business attached.

Its best idea is also its most portable: lower the blast radius of the first yes. That works when the sidecar has a real audience, separate economics and senior sponsorship - and when success criteria are agreed before launch. It does not work when the pilot becomes a permanent orphan, the institution lacks integration discipline or executives mistake a quick front end for a safe conversion. Banks do not need more confidence theater. They need evidence, escape hatches and a partner willing to remain in the room when the queue hits 74,584.