The Great Infrastructure Shift - Payments 4.0 Summer School - Week 1
From Monoliths to Modular: Why Your Payments Stack Will Be Obsolete by 2026
Welcome to Week 1 of Payments 4.0 Summer School.
Over the next six weeks, we're going to dissect the most fundamental transformation in payments since the internet. Instead of just writing about the trends, this newsletter series is about taking seemingly different topics and piecing them together, so we as Payments Professionals get a better understanding of what is happening in our industry, how it is affecting us, and what our strategic choices are.
Let’s Begin…
Zelle went dark for four hours on May 2nd. Bank of America, Truist, and Navy Federal customers found their rent payments suspended in digital limbo. "I feel like my money is suspended and nobody will do anything or take any blame!" one user complained on DownDetector.
Meanwhile, Cybersource has been wrestling with Mastercard authorization failures since July 10th. Global Payments suffered transaction processing issues on June 27th. Payment failures now kill growth for 94% of app businesses within 12 months, according to new research from Primer.
Legacy payments infrastructure isn't just slow anymore. It's breaking.
While legacy payments companies patch and upgrade their monolithic systems, a parallel infrastructure has emerged. Stablecoins now facilitate $20 billion to $30 billion of real payment transactions daily. Their on-chain volume reached $5.6 trillion last year, accounting for 35% of Visa's total transaction volume. The USDC market cap doubled. PayPal's PYUSD is settling merchant transactions. Mastercard announced support for four stablecoins on its Multi-Token Network.
The transformation isn't coming. It's here.
Twenty Years of Accumulated Technical Debt
The current payments stack across most incumbents was built for a different world.
Companies like Fiserv, Global Payments, and others, now among the world's largest payment processors, run on foundations laid in the 1970s, 80s, and 90s through acquisitions of legacy providers.
Consider the architecture beneath today's giants.
Fiserv's empire was built through acquiring companies like CheckFree (which itself had absorbed Corillian and Carreker), plus countless smaller processors with decades-old codebases. Global Payments emerged from the 2019 merger with TSYS, a company founded in 1959 as a division of Columbus Bank and Trust. Before that merger, Global Payments had already absorbed Heartland Payment Systems in 2016, bringing along its collection of legacy POS systems and processing platforms.
These acquisitions created technological layer cakes.
Each company brought its own systems, built in different eras, using different programming languages, operating on different assumptions about transaction volumes and payment methods.
The result: platforms that process billions of transactions daily while running on architectural decisions made when credit cards were swiped through carbon paper.
The inadequacy shows most clearly when new payment technologies emerge.
Take SoftPOS (Software Point of Sale), which transforms NFC-enabled smartphones into contactless payment terminals. Modern cloud-native companies can implement SoftPOS solutions in three months and achieve PCI compliance in six months. AWS case studies show companies like SoéPay building complete SoftPOS platforms, including payment gateways, terminal management systems, and merchant portals, in just three months.
Legacy processors require 18-24 months for the same integration. They must thread new functionality through multiple acquired systems, update compliance across different platforms, and ensure compatibility with decades of accumulated code. When Apple announced Tap to Pay on iPhone, nimble processors had working integrations within weeks. Legacy giants are still working on theirs.
Stablecoins present the same challenge.
Circle launched USDC settlement capabilities that modern processors integrated rapidly. Traditional processors are still evaluating how blockchain settlement fits into systems designed when the fastest settlement option was next-day ACH.
Most monolithic infrastructures can't adapt fast enough. They weren't designed for AI-driven routing decisions or real-time stablecoin settlement. Think Steam engine trying to power a jet aircraft.
But this brings us to a critical realization: the same forces that make legacy systems slow to adapt are creating opportunities for fundamentally different approaches. This sets the stage for what the industry calls "mesh architecture," but let's first understand what that actually means.
The Mesh Revolution: Intelligence at Every Layer
Mesh architecture (also called modular architecture) represents a complete departure from monolithic thinking. Instead of one system handling all functions, mesh breaks payment processing into independent, specialized services that communicate through APIs.
The mesh model dismantles the monolith entirely.
Each function exposes a stateless API: token vault, 3-DS authentication, fraud detection, and network gateways. An orchestration layer stitches them together in real time, creating a decision engine that optimizes every transaction.
This isn't just an architectural change. It is how competitive advantage will be created in the future.
Modern orchestration platforms inspect transactions in under 40 milliseconds, then route to the rail with the lowest total cost while meeting risk requirements. They cache scheme rules to route U.S. debit to the cheapest network or trigger pay-by-bank flows when card interchange outweighs conversion drag. They log granular metadata for continuous machine learning optimization.
The results are measurable.
Merchants switching from one-size-fits-all routing to intelligent meshes reduce processing expenses by 10-20% while improving authorization rates. Retries happen in milliseconds, not minutes. Failed transactions get routed to backup acquirers before customers notice.
Three Forces Making Mesh Inevitable
However, modular infrastructure has been around for quite a while; the reason it now seems to hit its stride is that three new forces are making it inevitable.
Open Banking APIs: U.K. and EU regulations forced banks to publish payment-initiation APIs, creating account-to-account alternatives to cards. Brazil's Pix processes 76% of all transactions. India's UPI handles 16 billion monthly transfers. Both displaced cash and cut merchant costs to cents per transaction.
ISO 20022 Messaging: This global standard provides richer data across payment rails. Orchestration layers no longer need bespoke adapters for each instant payment system. The same API call can route to SEPA Instant, FedNow, or RTP depending on cost and speed optimization.
Cloud Infrastructure: Containerized runtimes let mid-tier PSPs spin up global routing logic without capital expenditure. What required data center investments and months of deployment now happens in weeks. Cloud-native processors like Silverflow enable direct connections to Visa and Mastercard through AWS infrastructure with 600+ global points of presence.
This architectural shift creates natural migration paths from monolithic systems to more flexible, responsive infrastructure. It also enables the next layer of innovation: new settlement mechanisms that weren't possible with traditional architecture.
Stablecoins: The New Settlement Layer
Understanding mesh architecture helps explain why stablecoins represent more than just another payment method. They're becoming the settlement infrastructure that makes real-time, global, programmable payments possible.
Most payments professionals see stablecoins as another payment method to support.
That's wrong.
Stablecoins are becoming the infrastructure layer that rebuilds payments from the ground up.
48% of enterprises cite real-time settlement as stablecoins' primary advantage. Lower fees come last. Speed beats savings because speed enables entirely new business models.
Visa has settled more than $225 million in stablecoin volume through participating clients. Not pilot programs. Production settlement. JPMorgan's JPM Coin clears $1 billion of internal transfers daily (now that their CEO Jamie Dimon has changed is opinion about crypto, this will only increase). Circle's USDC enables 24/7 cross-border settlement without correspondent banking delays.
All of this creates strategic implications that will compound:
Liquidity Management: Instead of prefunding accounts across multiple markets, treasurers will hold USDC and convert to local currency on demand. No correspondent bank fees. No multi-day settlement delays. No foreign exchange margin stacking.
Programmable Money: Smart contracts will enable automatic settlements, conditional payments, and complex multi-party transactions. AI agents can execute payments autonomously using predefined logic. When software makes payment decisions in microseconds, your infrastructure either keeps up or gets bypassed.
Network Effects: Mastercard already enables four stablecoins on its Multi-Token Network, and Visa believes every institution that moves money needs a stablecoin strategy. In other words, the card networks aren't fighting against blockchain settlement; they're building on it.
The Enterprise Adoption Wave
Wall Street firms already use USDC to settle tokenized money market funds, unlocking 24/7 transaction finality while bypassing legacy clearing systems. For companies such as Franklin Templeton and BlackRock, this isn’t just an experiment anymore; it’s their regulatory operations.
JPMorgan built proprietary infrastructure for stablecoin settlement. Their JPMD token allows round-the-clock settlement for institutional clients seeking faster, cheaper transactions while maintaining traditional banking connections.
This isn't crypto companies disrupting finance; this is finance rebuilding itself on crypto infrastructure.
This transformation of settlement infrastructure requires technical capabilities that mesh architecture enables, but monolithic systems struggle to provide. Which brings us to the foundation that makes both mesh payments and stablecoin settlement possible: API-first design.
API-First Architecture: The Technical Foundation
The shift from monoliths to mesh requires API-first architecture from day one.
This enables real-time integration where updates deploy continuously instead of quarterly. New payment methods, fraud rules, or routing logic are implemented in days, not months.
Stripe provides a perfect example.
Their comprehensive API suite handles everything from basic card transactions to complex marketplace payments, subscription billing, and global tax compliance. They support 135+ currencies and dozens of payment methods with machine learning systems that dynamically adjust payment flows based on risk profiles.
The modular approach creates elastic scaling.
Companies like Postmates use Square's APIs to handle complex, high-volume transactions seamlessly. When transaction volumes spike, the infrastructure scales automatically. When new payment methods emerge, they integrate through standardized APIs without rebuilding core systems.
The Embedded Finance Multiplier
API-first architecture doesn't just improve existing payments. It enables entirely new business models where financial services embed directly into software platforms.
Shopify recognized this early.
Rather than relying on traditional payment providers, they partnered with Stripe to create embedded payment experiences that keep merchants within their ecosystem. Shopify Payments now processes billions in transactions. Shopify became an e-commerce giant that controls the financial layer of merchant businesses.
HubSpot, Xero, and other B2B SaaS platforms integrated payments into core workflows, eliminating external payment processors. Payments became seamless features, not separate services. Customer retention increased. Revenue per customer grew. Platform stickiness improved.
The platforms embedding payments win. The companies enabling embedded payments win. Traditional payment providers watching from the sidelines lose market share to more integrated competitors.
This API-first foundation enables rapid adaptation to new technologies, but it also creates new categories of risk that didn't exist in monolithic systems. The distributed nature of mesh architecture means that security and operational challenges manifest differently.
Risk Redistribution in Modular Systems
Modular architecture redistributes risk rather than eliminating it. Each API adds dependencies. Each integration creates potential failure points. Attackers adapt to exploit new vulnerabilities.
PCI DSS 4.0 responds by tightening multi-factor authentication requirements and demanding continuous monitoring across every service provider in the mesh. DORA in the EU extends operational resilience duties to critical third parties. An orchestration vendor outage can trigger regulatory fines at the merchant level.
MiCA forces any stablecoin rail touching Europe to attach travel-rule metadata, pushing compliance tasks into payment gateways. PSD3 drafts suggest stricter customer confirmation windows that instant payment rails must meet without sacrificing user experience.
The positive side effect: cybersecurity innovation accelerates to meet these challenges.
Passkey adoption doubled in 2024, with Japanese wallets reporting near-zero phishing fraud after switching from one-time passwords. Modular stacks make it easier to plug in newer security innovations and swap them out when stronger solutions appear.
These risk considerations directly impact how companies must restructure their operations, partnerships, and business models to compete effectively in the modular payments landscape.
Strategic Consequences: Four Business Model Shifts
The infrastructure transformation forces strategic changes across four dimensions:
Technology Operations
Engineering teams shift from quarterly releases to weekly rollouts.
Observability, latency, authorization lift, and cost per rail become product metrics instead of operational dashboards. Traditional financial institutions take 20-30 months to launch new products due to legacy system limitations. Next-generation platforms reduce this to weeks through cloud-native, API-first architectures.
Partnership Strategy
Sourcing transforms from single-vendor RFPs to portfolio management.
Companies dual-source acquirers and stablecoin issuers like banks, multi-home cloud providers. Vendor scorecards weigh switching friction as heavily as feature depth. The ability to replace underperforming providers without rebuilding core systems becomes strategically critical.
Commercial Models
When intelligent routing cuts blended fees by double digits, pricing negotiations center on residual value: fraud detection accuracy, data enrichment quality, and authorization optimization effectiveness. PSPs unable to prove differential performance watch market share shift to cheaper, more effective alternatives.
Treasury Management
Instant rails and blockchain settlement compress days-sales-outstanding.
Marketplaces offering weekend payouts use USDC to unlock working capital cycles that competitors stuck in T+2 card funding cannot match. Liquidity risk migrates from batch funding models to real-time balance sheet exposure, requiring hourly treasury management.
Understanding these business model shifts helps clarify the strategic options available to different types of companies as the industry transforms.
Three Strategic Paths Forward
Every payments company faces three strategic options:
1. Orchestrators: Become the Neutral Meta-Rail
Orchestrators win by abstracting complexity.
The strategy involves deepening decision intelligence: ingesting authorization codes and issuer data, feeding machine learning models, surfacing "next-best-rail" scores merchants can act on immediately.
Next-generation orchestrators optimize beyond cost.
They route EU consumer payments to MiCA-approved stablecoins when advantageous, fall back to SEPA when blockchain liquidity is thin, and comply with regulatory requirements automatically through configuration rather than custom development.
2. PSPs: Build Rail-Agnostic Checkout Primitives
PSPs control the merchant user experience.
Their strategy involves exposing single "/pay" endpoints that execute card, bank, wallet, or tokenized deposit flows transparently.
Success requires:
Schema-flexible token vaults supporting PANs, IBANs, and wallet addresses
Sub-100ms fraud scoring APIs operating without browser context
Pricing models sharing savings from account-to-account or stablecoin transactions with merchants, aligning adoption incentives
3. Acquirers: Differentiate on Settlement Speed
Acquirers historically differentiated on BIN reach.
Now they must differentiate on liquidity. The strategy involves connecting stablecoin rails and instant payment corridors directly to funding engines, offering hourly or event-driven merchant payouts.
The acquirer integrating blockchain settlement while providing fiat net-settlement finality in unified dashboards will capture global marketplace flows frustrated with traditional cross-border cutoff times.
Critical Architecture Questions
So, how do you take this new understanding and implement it yourself?
Companies evaluating current infrastructure against Payments 4.0 reality should answer these questions honestly:
Infrastructure Readiness
Can legacy systems route transactions to USDC settlement in under 100ms? Or are they locked into traditional banking rails with multi-day settlement windows?
Do most providers have granular transaction data for machine learning optimization? Or are they limited to summary reports that provide insufficient detail for competitive decision-making?
Can incumbent platforms implement new payment methods in days? Or do they need months of development cycles when regulations change or new opportunities emerge?
Competitive Positioning
What happens when merchants bypass traditional providers entirely? Enterprise merchants already connect directly to card networks through cloud-native processors. Legacy providers must demonstrate sufficient value to remain relevant.
How do traditional systems compete when AI agents choose payment rails? When software makes routing decisions based on microsecond analysis of cost and authorization probability, legacy infrastructure struggles to keep pace.
Can established players adapt faster than legacy systems allow? When stablecoin regulations clarify or new payment methods gain adoption, incumbent response times often lag behind market opportunities.
Take Action
Companies that decide to make strategic moves now will define competitive dynamics for the next decade.
So what should you do?
If working at a traditional payments provider, building API-first capabilities becomes urgent. Monolithic infrastructure has 18 months of competitive viability remaining. Smart incumbents are planning modular architecture transitions to avoid becoming backend providers with shrinking margins.
If running an enterprise merchant processing over $500 million annually, direct-to-scheme processing (with a dedicated BIN from an Acquirer) deserves serious evaluation. The economics are compelling, but strategic advantages around control and data access provide more lasting value. Forward-thinking companies are reducing dependence on traditional acquirers before they become commodity providers.
If operating a software platform, embedded payments aren't optional anymore. Customers expect financial services to be native to product experiences. Successful platforms partner with API-first providers rather than losing customers to competitors offering more complete solutions.
If building infrastructure, focusing on the decision layer beats just moving transaction pipes. The future belongs to platforms making intelligent routing decisions, not merely moving money between endpoints.
The Infrastructure Reset
The shift from monoliths to modular architecture isn't incremental. It's a platform reset collapsing costs, widening choices, and reframing risk management.
Stablecoins aren't additional payment methods. They're becoming the settlement layer for next-generation financial infrastructure.
API-first architecture isn't a technical preference. It's the foundation enabling everything else.
Companies re-architecting around decision engines rather than processors will route every transaction to optimal rails, fund merchants in minutes, and absorb regulatory changes through configuration updates instead of re-platform projects.
Companies clinging to monolithic architectures will discover that bots, not shoppers, have routed around them entirely.
The question isn't whether this transformation happens. It's whether companies lead it or spend the next decade catching up with competitors who moved first.
Next week, we examine how merchants evolve from simple payment acceptance to full-stack financial platforms. We'll explore the four-stage maturity curve separating reactive players from strategic winners in the embedded finance economy.
OPTIONAL: For homework, audit the current infrastructure against the three architecture questions above. Companies facing reality first will shape what comes next.
Thank You for Reading. Please like, Comment, Share, or Post on Your Social media. I appreciate all the feedback I can get.
P.S. If you want to work with me in a larger capacity, either speaking, advisory, or consulting, feel free to email or DM me.