The modernization of payments isn’t happening in a single place or in a single way. It’s happening everywhere and almost always in the cloud. For Amazon Web Services (AWS) and its financial-services partners, the move to cloud infrastructure is more than an IT project. It’s a structural shift redefining how money moves, how institutions manage risk and compliance and how value is created in a digital economy that never sleeps.
Across global banks, payment networks and FinTechs, the story is the same: real-time expectations and regulatory resilience are forcing institutions to rebuild from the ground up. In the United Kingdom, NatWest is re-architecting its payments stack on AWS to deliver instant cross-border transactions and continuous reliability. In the United States, the Federal Reserve’s FedNow® Service is using cloud-native design to bring instant payments to more than 1,400 financial institutions—a national experiment in real-time trust. And in global business payments, Convera is using AWS to turn cloud migration into a platform for data-driven intelligence and agentic automation across over 200 countries and territories and more than 140 currencies.
Each represents a different stage in the same evolution: from modernization to intelligence, from speed to value. As AWS executives Nilesh Dusane and Vishal Arora note throughout this eBook, the power of the cloud lies not only in scale, but in what happens after the move—when data, AI and microservices combine to deliver faster, safer, more personalized financial experiences.
In the new payments landscape, modernization isn’t just about moving to the cloud. It’s about what you build and achieve once you’re there.
For global banks, instant payments and digital-first experiences are forcing a rethink of core infrastructure long weighed down by legacy systems. The question is how to make those systems flexible enough to support business lines while meeting the regulatory and resilience demands of real-time transactions.
That’s where cloud re-architecture comes in. Over the past few years, the industry has moved beyond shifting applications to the cloud to rebuilding them as cloud-native systems.
Nilesh Dusane, global head of institutional payments at AWS, said the change marks a turning point. “Five or six years ago, customers started their cloud journey with what we called a lift-and-shift model,” he said. “You could scale and cut costs, but now customers like NatWest are re-architecting their payment applications using cloud-native services.”
Breaking down monolithic systems into APIs and microservices lets banks focus on core strengths while gaining flexibility.
For NatWest, this meant aligning its modernization push with its broader cloud strategy. “I’d stress the willingness and preparedness to be bold on this journey,” said Ian Povey, NatWest’s Chief Information Officer (CIO) of Payments Technology. “Without that ethos, the challenges could be insurmountable.”
AWS teams, Povey added, helped align the project with regulatory expectations. “We have specialists who know how to speak to regulators, who know what they are looking for,” Dusane said.
NatWest also worked with Icon Solutions, which provided its low-code Icon Payment Framework (IPF) on AWS. The partnership helped NatWest’s teams roll out Single Euro Payments Area (SEPA) instant payments and Clearing House Automated Payment System (CHAPS) on AWS, operating as a single collective. The joint steering committee includes AWS, Icon and NatWest representatives, ensuring transparency and accountability.
The collaboration paid off: NatWest went live with a multi-region solution in just over four months, meeting regulatory obligations and advancing its goal of preparing for “multiple schemes becoming instant or near-instant,” Povey said.
From APIs to AI: Building the Next Generation of Payments
A major outcome of the modernization is a culture of collaboration and experimentation. By embedding DevOps engineers into site-reliability teams, NatWest created what Povey called “cross-fertilization” between development and operations. “We actually reversed the polarity,” he said, “embedding DevOps into site reliability engineering (SRE) and SRE into DevOps for a deeper understanding of both sides.”
The move toward APIs has accelerated the bank’s ability to integrate new payment rails. Povey, who previously worked at PayPal, said APIs unlock flexibility: “We’re catering for multiple different payment streams. For me, it’s an abstraction on all dimensions, not just the front end.”
For AWS, the next phase centers on “time to value” rather than “time to market.” Dusane said banks now want their customers—retail, commercial and corporate—to quickly benefit from the new systems they’ve built on AWS. APIs are critical to that goal.
NatWest’s modernization is also opening the door to AI and machine learning (ML). Povey’s team is testing AI tools that simulate transaction “war games” and automate quality-assurance tests. “AI is becoming a powerful tool for reverse-engineering legacy systems,” he said, helping uncover what should be rebuilt or retired.
Dusane said AWS is now seeing banks use generative AI to boost developer productivity, enhance employee efficiency and create hyper-personalized services. A relationship manager, he noted, could use AI to draft 80% or 90% of a client proposal before tailoring it further.
As banks adopt ISO 20022 standards, the benefits go beyond compliance. Richer data can improve fraud detection, deepen customer insights and enable value-added services built on top of core payment capabilities.
“Payments is a network sport,” Povey said. “It requires a different level of collaboration—not just among banks and partners, but with customers—to get to the right use cases.”
Instant payments are redefining how money moves—and how financial institutions meet rising expectations for speed and reliability. The Federal Reserve’s FedNow® Service, launched in 2023, now counts more than 1,400 participants, up from just 35 at launch. Transaction volumes have grown by more than tenfold year over year, FedNow Executive Vice President and CIO Dan Anthony said.
The growth isn’t just about numbers. “One of the biggest milestones that makes me realize the impact of instant payments was processing FEMA disaster relief through FedNow,” Anthony said. “That’s a mark we’ve really arrived.”
A key to that success lies in FedNow’s cloud-native design—a decision made during the pandemic that has proven critical. “Launching in the cloud affords us the ability to move fast and to innovate,” Anthony said. Traditional physical provisioning “would have been a nightmare.” Instead, the service can scale quickly and onboard institutions in days. “Our fastest onboarding from digital signature to go-live was seven days,” he noted.
That speed is encouraging banks to modernize. Dusane noted that FedNow has become a catalyst for financial services customers. “By taking instant payments, banks have started to modernize their payment stack using cloud-native services,” he said. “That enables faster time to market and hyper-personalized services for retail and commercial customers alike.”
The shift, Dusane added, moves institutions toward true cloud-native design. Some banks are already using microservices to enable dynamic payment routing, so a payroll missed on ACH, for example, can be instantly rerouted through FedNow. Others are building “just-in-time expense reimbursement,” pushing funds to employees as soon as an expense report is approved.
Anthony called that the kind of innovation FedNow was meant to inspire. “That’s exactly the use case I’d love to see more of,” he said.
From Speed to Trust: Building the Next Payment Standard
FedNow’s rapid adoption has pushed the Federal Reserve to keep expanding transaction limits and services. “We launched with $500,000, went to $1 million earlier this year, and almost immediately saw maximum-value transactions flow through,” Anthony said. “There’s a lot of appetite for $10 million transactions.” The Fed has also added new risk-management tools and automation features, along with account-activity thresholds.
Anthony said requests for payment (RFPs) could be transformative for B2B use. “You can start to orchestrate business processes—invoicing and receipt of payment—in a fully automated way with real-time payments coming through,” he said.
For banks, however, instant rails are only part of the equation. “FedNow can do instant payments,” Dusane said, “but if the processes before that are batch, then you’re not giving customers the full benefit.” True real-time capability, he added, demands 24/7 ledger access, near-instant fraud checks and embedded KYC verification.
Building that infrastructure requires experimentation. “Technology is the easy part,” Dusane said. “The hard part is finding business value in how you use FedNow.”
Cloud delivery, Anthony said, lets FedNow bake in resilience and security from the start—key ingredients for customer trust. “Resilience isn’t an add-on,” he said. “It’s built into the product.”
Both executives pointed to data as the next frontier. The ISO 20022 standard, Anthony said, offers “a couple thousand fields” of structured information that banks and corporates can use to automate reconciliation, improve forecasting and fight fraud. “With instant payments and the cloud, the industry will get access to richer data—and that will enable innovation,” Dusane said.
Anthony summarized the vision simply: “Instant payments reach their full potential when banks can both receive and send them—managing finances with full flexibility and, to paraphrase, moving money at the speed of need.”
For payments companies, moving to the cloud isn’t the finish line—it’s where the real work begins.
When global business payments giant Convera became an independent company, it faced the challenge of transforming an infrastructure designed for a slower era of finance. Its network spans more than 200 countries and territories, and over 140 currencies, but its systems were still anchored to on-premises hardware.
“We are a carve-out business and very standalone today,” said Sudipto Das, vice president of engineering at Convera. “It was an incredible 10 months to move everything from on-prem to the cloud. But the real story starts after that—that’s when you start unlocking all the benefits.”
That shift, done with AWS, became a foundation for innovation, altering how data, compliance and intelligence interact across global markets.
“Moving cross-border B2B payments infrastructure to AWS delivers immediate modernization gains,” said Vishal Arora, head of generative AI and machine learning for payments at AWS. “It helps address the core challenges traditional payment systems face.”
For Convera, the payoffs came quickly: scalability, global reach and faster innovation cycles. Legacy infrastructure required provisioning for predictable month-end spikes, but cloud computing removed that constraint.
The cloud also reshaped Convera’s approach to compliance and data management. “A thorough understanding of all data types and flows is essential,” Arora said. Customers are increasingly using automated discovery tools to classify and control sensitive information through encryption, tokenization and geofencing so “regulated data never leaves its required boundaries,” he said.
Das said AWS handles the physical security and certifications, while Convera manages access control, network segmentation and encryption. Automation and observability have turned compliance from an annual audit into a continuous process. “It’s about amplifying customer trust,” he said. “AWS gives us a secure foundation—and we build the intelligent, compliant, customer-centric layer on top.”
From Cloud to Intelligence: The Agentic Future of Payments
Once the data is in order, Convera’s focus turns to intelligence. “Our customers are leveraging generative AI to improve productivity, reduce fraud and deliver more tailored experiences,” Arora said. Adoption is growing across five areas: customer support, document processing, payment operations, fraud detection and hyper-personalization.
Das said Convera invested heavily in organizing its data before diving into AI. “Only then did we build models for FX routing and treasury management,” he said. “Quality inputs lead to quality outputs. Better user stories accelerate engineering cycles.”
Both executives see the next wave coming from agentic AI—systems that act autonomously within defined boundaries. AWS supports what Arora calls “agentic capabilities” through its Amazon Bedrock AgentCore, a managed service for deploying and monitoring AI agents securely at scale.
No modernization story is complete without tackling stablecoins and tokenized settlements. Das sees them as complementary, not yet replacements, for traditional rails. “Unless the entire world moves into the stablecoin world, there will still be a need for off-ramping into fiat currencies,” he said. Companies can play different roles—accepting stablecoin payments, using token rails for specific corridors or focusing on last-mile conversion — depending on regulatory maturity and interoperability.
Arora defined success metrics: 24/7 cross-border transfers that settle in seconds, reduced costs, freed liquidity and built-in smart-contract compliance. Transparency and auditability, he added, will be essential for trust and oversight.
For both AWS and Convera, modernization is as much a mindset as migration. “The new currency of global payments,” Das said, “is not just data or speed—it’s trust.”
