Build vs. Buy: How FX Payments Fintechs Should Decide What to Build

If you run an FX payments fintech, you've probably asked whether to build your own platform. AI coding tools make it look easier than ever: a working prototype can come together in weeks, and building can feel like the obvious move.

We've spent 12 years building software in FX and cross-border payments, and our view is clear: for most FX payments fintechs, building your own platform is the wrong decision. It gets you started "cheaper" and then costs you more, in money, time, and risk, as the business grows. Here's why, and how to decide when building it yourself does make sense.

Key takeaways

• Building only looks cheaper at first. The real costs come later, in rebuilds, audits, and engineering time pulled away from growth.
• When you build, you solve every problem on your own, without an expert partner or ready-made tools for the work every regulated FX business has to do.
• AI speeds up the first version, but running a regulated platform still takes people, expertise, and ongoing spend. In the end, it will cost more than working with a partner.
• White-labeling the client experience on a platform you run yourself keeps the margin and the client relationship with you. Launching on someone else's payments license doesn't.
• Build the parts your clients choose you for. Buy the rest from a partner whose platform is turnkey for regulated businesses.

Why building looks right at the start

I've had this conversation many times, and the case for building usually sounds reasonable. You know your business better than any vendor does, your engineers are good, and a working demo is only a few weeks away. The reasons to build are real:

• It looks cheaper. A small team, a few integrations, a portal, and you're live.
• It fits exactly what you do today. No configuration and no compromises.
• You own the IP, which can matter to investors.
• AI makes it look faster still. A prototype is no longer a big investment.

For a narrow product in its first months, some of this holds. The trouble is that a regulated FX business doesn't stay narrow for long.

What building actually costs

The costs of building rarely show up in the first budget. They arrive later, one at a time, and usually when the business is growing fastest. Here's where they tend to land:

• You work everything out by yourself. There's no expert partner and no playbook for onboarding, reconciliation, security, safeguarding, or regulatory reporting. You find out what "right" looks like from an auditor, a regulator, or a banking partner, after you've shipped.
• You rebuild what's already solved. Client onboarding and KYC, multi-currency ledgers, reconciliation, trade capture, confirmations, statements, and regulatory reports are commodities. None of them is where an FX business wins clients, and every one of them has to be right.
• Time and money go to plumbing instead of growing the business. The engineers who could be opening new corridors, launching products, and serving clients spend their time keeping the platform running.
• Growth exposes what you didn't know. More volume, forwards, multi-currency accounts, a new corridor, or a new regulator each test the platform. Many fintechs reach a point where they realize much of it needs to be rebuilt, usually at the worst possible time.
• Regulation keeps moving. New reporting rules, audits, and banking-partner due diligence all land on your own engineering team. When FINTRAC overhauled how electronic funds transfers are reported in 2024, we updated our EFT reporting to the new rules, connected to FINTRAC's new APIs, and helped every client migrate, at no extra cost to them.
• Key-person risk. A home-grown platform often lives in the heads of two or three people.

Treasury shows how much precision this takes. Managing liquidity, risk, and settlement across currencies depends on accurate data and tight controls, which we covered in Managing Treasury in Regulated Payments Businesses.

AI changes the build, not the ownership

We use AI throughout our own engineering, and it has changed how we work (see How Kooltra Builds, Part 2: Our AI-Assisted Engineering Workflow). It makes the first version of anything faster to write.

It doesn't change what it takes to run a regulated FX business: knowing what a regulator or banking partner will ask for, reconciling every currency every day, and keeping an audit trail that holds up. Building that properly still takes people's time, deep domain expertise, and a real AI bill. Added up, it costs more and carries more risk than working with an expert partner from day one. None of this is an argument against AI. It's an argument for aiming it at the work that enhances your competitive advantage.

What about a white-label?

"White-label" can mean two very different things, and the difference matters. One is launching on another provider's platform under its payments license and banking relationships. It's a quick way into a market, but the products, pricing, and client relationships stay with the provider, which is why many fintechs move on once they hold their own payments license.

The other is white-labeling the client experience on a platform you run yourself, and it's the one we offer. With Kooltra, your clients see your brand across the client portal, onboarding, and documents like trade confirmations and statements, while you keep your own license, pricing, and client relationships. And if you provide liquidity or payments to partners of your own, you can give each partner its own branded experience on your Kooltra platform.

How to decide

You don't have to build or buy the whole platform. The fintechs I've seen get this right buy the foundation every regulated FX business needs, and build only the pieces their clients actually choose them for. Here's where I'd draw the line:

Build what sets you apart: pricing logic, a distinctive client experience, a niche corridor, or proprietary risk models.

Buy what every regulated FX business needs: onboarding and compliance workflows, the ledger and accounts, reconciliation, reporting, and treasury.

A simple test: if a client would never notice the difference between your version and a well-run alternative, it's a commodity. Buy it, and put your engineers on the parts they would notice.

Before you commit engineering time, answer these honestly with your team:

1. What will it cost to own this in three years, not to ship it in three months?
2. Who on the team knows what a regulator or banking partner will ask for?
3. What happens when we add a new product, a new currency, or a new country?
4. Which parts of this are truly ours, and which would any regulated FX business need?
5. If we build and have to move later, what will that cost?

If you've already built

Signs it's time to move: a rebuild is coming, audit findings keep landing on engineering, most of the team's time goes to maintenance, or a new product is blocked by the platform.

Moving doesn't mean starting over. Your clients, data, and workflows come with you, and your engineers go back to working on what's yours. We're helping a growing cross-border payments fintech make exactly this move right now, off the platform it wrote itself and onto Kooltra. Our implementation team has run these migrations many times, and they'll work alongside yours from the first requirements session to the day you switch the old system off.

How Kooltra helps

We handle the compliance, operations, and reporting work every regulated FX business needs, so your team can spend its time on your clients and your growth. In practice, that means:

• Ready for regulated businesses from day one. Onboarding, compliance, trading, payments, reconciliation, and reporting workflows are already in place, so you start with what a regulator and banking partner expect to see.
• 12 years in FX and cross-border payments, and more than $500 billion in annual FX processed for our clients.
• A team that has run FX businesses, not just written software for them. Kooltra started inside an FX trading operation.
• Configurable workflows that business users can change without waiting on engineering.
• Modern APIs to extend the platform and build your own client experience on top of it.
• New businesses go live in weeks, not the months a build takes.

You keep building what wins you clients. Kooltra runs the rest.

Banks and money services businesses face their own versions of this decision, and the answer looks different for each. We'll cover them in the coming months.

Book a strategy call to see how Kooltra can support your business.