AI has fundamentally changed how software gets built, and the pace of that change is only accelerating. What used to take months now takes weeks, and in many cases, what once took weeks is now happening in days. Developer productivity is rising sharply as AI coding tools reduce the effort required to build, test and deploy new features, allowing teams to iterate in near real time rather than on rigid release cycles.
Kevin Kidd, founder and CEO of Cresora Commerce, argues that payments
infrastructure, not code, now sets the pace of innovation in AI-driven software.
This shift is not just improving efficiency. It is resetting expectations across the entire market. Software companies are no longer optimising for static roadmaps. They are building adaptive systems that evolve continuously based on customer needs, operational realities and competitive pressure. As a result, their partners are now expected to operate with the same level of speed and flexibility.
That is where the problem starts.
Payments has not kept pace
Payments infrastructure has not kept up with this new pace of software development. For years, the industry has operated under a simple assumption: once payments are embedded, the problem is solved. A provider is selected, APIs are integrated and the system is considered complete. That model worked in a world where software evolved slowly and changes were infrequent.
It does not hold up in an AI-driven environment where change is constant.
Modern software is no longer static. Workflows shift, pricing models evolve and customer experiences are continuously refined. At the same time, more companies are building vertically focused software tailored to their specific industries rather than relying on generic third-party platforms. AI has lowered the barrier to entry, making it easier to create highly customised solutions that align closely with how a business actually operates.
However, building the front-end experience is only part of the equation. The real complexity sits beneath it.
Payments, settlement, reconciliation and financial reporting are deeply interconnected workflows that span multiple systems and stakeholders. These processes are not modular in the way most APIs suggest. They are tightly coupled to operational logic, compliance requirements and financial controls. When infrastructure is rigid, every change to the product layer introduces friction at the financial layer.
Where legacy providers struggle
Many established financial services platforms are built on older code bases and layered integrations that were never designed for rapid iteration. Customisation often requires significant engineering effort, and even small changes can introduce risk across reconciliation, reporting and compliance workflows. In environments where accuracy and auditability are critical, that risk slows everything down.
The result is a growing disconnect between how fast companies can build and how fast their infrastructure can adapt.
Over the next one to two years, this gap will become much more visible. As AI continues to compress development timelines, the limiting factor for innovation will not be engineering capacity. It will be the ability of underlying systems to support continuous change without introducing operational instability.
This is already starting to show up in how companies think about their technology stack. Increasingly, organisations are questioning whether their payment infrastructure is enabling growth or quietly constraining it. Many recognise that switching providers is difficult, not because better options do not exist, but because the cost and risk of unwinding deeply embedded systems is too high.
At the same time, the demands on these systems are increasing. Businesses are operating across more channels, more geographies and more complex financial models than ever before. They require real-time visibility into transaction flows, faster financial close cycles and greater flexibility in how they route and manage payments. Yet many are still relying on infrastructure that was designed for a much simpler operating environment.
A structural bottleneck
The industry has spent the last decade making it easier to embed payments. That was an important step, but it solved for a different era of software. The challenge now is not embedding functionality. It is enabling adaptability.
In a world where software can change overnight, infrastructure that takes quarters to adjust is no longer just inefficient. It actively limits what the business can do.
The companies that win in this next phase will not be defined by the payment provider they selected years ago. They will be defined by whether they have built the flexibility to change providers, workflows and financial configurations without needing to rebuild their systems each time.
Because in an AI-driven world, the real advantage is not speed alone. It is the ability to sustain that speed across every layer of the business.
Kevin Kidd is founder and CEO of Cresora Commerce, where he leads product strategy for AI-first commerce infrastructure that orchestrates payments, data and financial workflows across fragmented systems.
The post AI Software’s Real Bottleneck is Payments Infrastructure appeared first on The Fintech Times.
