Reimagining
Critical Systems

Modernize without disruption

Why AWS

AWS Transform uses agentic AI to analyze, document, refactor, and migrate mainframe applications, compressing timelines from years to months
Operational continuity maintained throughout: modernize mission-critical systems while they
remain live
One of the broadest FSI partner networks for migration from modernization specialists to
next-generation core providers
Unified, AI-ready data foundations that replace fragmented legacy data stores with open architectures at scale
18 years of financial services migration expertise, with proven results

Modernize without disruption

Legacy systems are no longer just a cost burden – they're a strategic liability. Mainframe talent is retiring, maintenance costs are escalating, and rigid architectures block institutions from deploying AI or delivering the experiences customers demand. The question has shifted from "why modernize" to "how to modernize at speed" without disrupting operations.

"The need to be able to compose different experiences using microservices is driving the modernization of mainframe and core systems across the banking sector."

Charith Mendis,
Head of Banking Market Development, AWS

Legacy is the bottleneck

Insurance reimagined
in the cloud
How insurers are modernizing core systems to unlock data and enable AI across claims, underwriting, and distribution

Banking on the cloud
How banks are modernizing legacy systems, streamlining processes, and building adaptive platforms - with new tooling accelerating migration and application transformation

Tokenized money is here to stay
As stablecoins move from niche to board-level priority, banks must build the infrastructure now as technology choices may need revisiting within months

From concept to enterprise scale and the platform decisions
that matter
The institutions succeeding with AI invested in platforms and data first

The shifts unlocking speed, resilience, and AI-readiness in financial institutions. Explore the thinking behind these below.

Legacy core insurance systems for underwriting, policy administration, billing, claims, and document management were built as monolithic solutions, making them slow and difficult to adapt. Insurers are migrating to microservices-based core platforms in the cloud for the speed and agility they require. This playbook explores how insurers are modernizing mission-critical platforms with AWS to accelerate product development, improve resilience, and create the foundation that makes AI and data-driven services possible.

Insurance re-imagined in the cloud

"AWS not only offers us a safe and open environment to continue innovating, but it has also opened doors to the exploration of emerging technologies as promising as generative AI." Manuel Rodriguez, CIO, Agroseguro

Key takeaways

Legacy core systems are slow, expensive, and block new product development
Microservices-based architecture accelerates development cycles, improves scalability, and increases resilience
Cloud migration of policy administration systems is delivering 25% faster batch processing and freeing developers from maintenance
AI-assisted migration tooling is addressing API security vulnerabilities during cloud moves, boosting developer productivity by 40%

Why it matters

Many insurers still rely on decades-old core systems that limit agility, increase operational complexity, and slow innovation. At the same time, the "Great Retirement" era means insurers need to do more with fewer people – and modernization is critical to accomplishing both. A strong foundational architecture is what makes AI-powered features, new product types, and faster time-to-market possible at scale.

Legacy mainframe systems in banking often contain millions of lines of COBOL, PL/1, and Assembler code. Manual modernization has been extraordinarily time-consuming and error-prone, and extracting decades of embedded business rules was nearly impossible with previous tooling. The vision is longstanding but recent technological advances have brought it within reach. This report explores how banks are using AI-powered reverse and forward engineering, microservices decomposition, and cloud-native architecture to modernize core systems at a pace previously unattainable.

Banking on the cloud

"We're building a banking model that brings sophistication to everyone, not just the privileged few." Charith Mendis, Head, Banking Market Development, AWS

Key takeaways

AI-powered engineering reduces modernization times by up to 96%
AWS Transform and Kiro automate assessment, rule extraction, and dependency mapping
Microservices speed innovation and reduce maintenance overhead
New tools accelerate banking modernization and transformation

Why it matters

Banks that modernize infrastructure and operating models can respond faster to customer expectations, market shifts, and regulatory change. Those that delay risk higher operational complexity, slower innovation cycles, and rising technology costs. What once required years of manual effort – converting COBOL to Java, then decomposing monolithic Java into microservices – now happens in a single compressed phase.

Stablecoins and tokenized deposits have moved from niche product to board-level priority, and banks are now asking "how do we build this?" But the ecosystem is extremely large, fast-moving, and fragmented – with an array of Layer 1 and 2 blockchains, interoperability propositions, and regulatory-driven initiatives to navigate. This report examines the infrastructure decisions banks and payment providers must make now, the technology partners that can help future-proof their propositions, and why flexibility in architecture choices will be critical.

Tokenized money is here to stay

“We’ve been inundated with interest and inbound dialogue from a whole spectrum of TradFi players [this year]. They’re not just coming to us for solutions, but for counsel." Senior executive at one of the world’s largest cryptocurrency exchanges

Key takeaways

GENIUS Act and global frameworks have unlocked bank investment confidence
Cross-border and remittances are the clearest opportunity
Programmability – not just speed and cost – is the real differentiator
Build flexibly: technology choices may need revisiting within months

Why it matters

The technology and competitive ecosystem around tokenized money will continue to evolve rapidly. Banks choosing platforms and partners today may need to revisit those choices within months. Seamless integration into banking and payment workflows is critical to adoption and requires infrastructure investment, not just strategy.

Slalom: Commercial banking origination reset. Building intelligent, scalable platforms on AWS

IBM: Modernizing core banking and shaping an AI-first future for insurance with IBM and AWS

Harness: The blueprint for modern delivery. How Harness powers governance and scale in financial services

TEKsystems: Solving complex financial services challenges with modernization, data, and AI

NetApp: Modernizing enterprise storage in the cloud. How FSx for ONTAP delivers NetApp's power with AWS simplicity

Partner insights

AWS partners are at the forefront of financial services innovation, building solutions that help institutions modernize faster, deploy AI at scale, and stay ahead of an industry that never stands still. Explore what they're working on here.

HCLTech and Fiserv: Agentic AI-powered mainframe modernization – without the multi-year timeline

How Itaú modernized bank statements
The 50-year-old platform that now answers customer queries
in milliseconds

Transamerica: Grid optimization and unleashing agentic innovation
Halving infrastructure costs, then building agentic AI on top

Rocket fuels growth with unified customer data on AWS
A unified data foundation that drives conversion, recapture,
and growth

Gemini: Cut trading latency 80% with AWS Outposts and
Local Zones
Bringing institutional-grade low-latency trading to the cloud

Generali: Building a central data, analytics, and AI solution
with AWS
Scaling from 5 to 50 AI applications in three years

Legacy replaced, operations uninterrupted

How financial institutions are modernizing core systems, cutting costs, and unlocking new capabilities – without disrupting what's already running. Read real customer success stories below.

Danske Bank: Halving migration timelines with hyperautomation
Turning months of migration work into seconds

Moving 16,600 servers and 25 petabytes of data to the cloud is the kind of undertaking that can take years and slow a bank down in the process. Danske Bank decided it couldn't afford that. Using AWS, they built Lyn™ – a hyperautomation solution that turned months of manual migration work into a process measured in seconds, as part of their Forward '28 strategy to become a leading digital bank.

Danske Bank: Halving migration timelines with hyperautomation

"Implementing the hyperautomation solution at Danske Bank has been a game-changer, yielding significant savings in time and costs. These freed-up resources are now powering value-driven initiatives that directly benefit our customers."
Ashish Shekhar, Senior VP, Global Head of Technology Platforms, Danske Bank

Key takeaways

50% faster migration timeline achieved through automated tooling and custom runbooks
Cut dual-run costs by 50% by eliminating months of parallel infrastructure
Runbook creation dropped from 3-4 weeks to just 4 seconds
700+ employees trained on cloud best practices to sustain a culture of continuous innovation

Why it matters

Legacy infrastructure carries a hidden cost beyond the maintenance bill – time. For a bank the size of Danske, every hour spent keeping aging systems running is an hour not spent building what customers actually need. Hyperautomation changes that equation, freeing teams to move forward rather than stand still.

Read more

Conversion is a data problem rather than a funnel problem. By building a unified data foundation on AWS that connects, curates, and activates customer data, Rocket transformed how it understands and engages customers at scale. The results speak for themselves: a 20% increase in refinance pipeline, 10% conversion lift, and recapture rates three times the industry average.

Rocket fuels growth with unified customer data on AWS

"The real shift is from having data to understanding data – for the foundation of decisioning, AI, measurable business impact, and activation all across your organization in one single uniform way." Garima Sharma, Vice President of Engineering, Rocket

Key takeaways

Unified customer understanding beats isolated AI initiatives
Standardized architecture patterns are the fastest route to team velocity
Client 360 gave every business area a connected view across the full home ownership lifecycle
A strong data foundation absorbs acquisitions – Rocket onboarded Mr. Cooper and Redfin without starting over

Why it matters

Data without understanding has a ceiling. As AI initiatives multiply, unified and business-contextualized data becomes the deciding factor – for speed, for personalization, for competitive edge. In financial services, where trust and timing determine whether a customer converts, Rocket's journey shows what's possible when the foundation is right.

Complex infrastructure, regulatory pressure and the demand for speed don't have to pull in opposite directions. Transamerica proved that cost, quality and performance can all be achieved at once – cutting infrastructure costs by 50% through targeted chipset and architecture changes, then building repeatable agentic AI patterns across their portfolio. Optimize the foundation first, then scale intelligence on top of it.

Transamerica: Grid optimization and unleashing agentic AI innovation

"Traditional wisdom will tell you don't go after all three – cost, quality and performance. I'm here to tell you going after all three is actually possible."
Krishna Subramanian, Senior Leadership Actuarial Technology &
Finance, Transamerica

Key takeaways

The right combination of chipset and architecture halved Transamerica's infrastructure costs
AWS Fault Injection Service stress-tests resilience in production to find
real gaps
Agentic AI patterns are most powerful when built to one pattern, many use cases, applied at scale

Why it matters

Most organizations treat infrastructure optimization and AI adoption as separate programs. Transamerica ran them together and the infrastructure work made the AI work better. Cleaner architecture, lower latency, and embedded resilience created the conditions for agentic AI to deliver at scale.

Read more

Itaú – one of Latin America's largest banks, serving 70 million customers across 18 countries – was running mission-critical bank statements on legacy, monolithic infrastructure. Customers could only access three months of history, response times hit two seconds, and corporate accounts had no real-time reconciliation. After modernizing on AWS, Itaú now serves five years of history, handles eight million requests per month, and responds in 200 milliseconds.

Itaú: From legacy to intelligent – how Itaú modernized bank statements

"Modernization is a journey. This is not a one-time event. You need to keep evolving the platform every day." Alisson Carneiro, IT Executive, Itaú

Key takeaways

Historical data access expanded from three months to five years – with complex filtering and real-time reconciliation for corporate accounts
Amazon OpenSearch replaced relational and NoSQL databases entirely
Shuffle sharding and hedging mean outages affects the fewest customers
Intelligent search uses Amazon Bedrock to translate natural language into OpenSearch queries with 80% fewer search errors already

Why it matters

Itaú's legacy architecture was letting customers down at the moments that matter most. This modernization shows that financial institutions don't have to choose between stability and innovation. A strong foundational architecture – cell-based, resilient, cost-optimized – is what makes AI-powered features possible at scale.

By combining AWS Local Zones, Outposts, Direct Connect and VPC peering, Gemini cut trading latency from 8–15 milliseconds down to under 2, while reducing client onboarding from months to minutes. The result: institutional-grade crypto and predictions trading, accessible from the cloud, without sacrificing the speed or determinism that professional traders demand.

Gemini: Cut trading latency 80% with AWS Outposts & Local Zones

"The live trading environment built through our predictions marketplace
has been one of the best ways we've been able to validate the platform performance at scale and all that we've built with AWS."
Paget Stanco, VP of Institutional, Gemini

Key takeaways

Distance is latency – moving compute into the NYC Local Zone put Gemini's infrastructure nearly co-located with its matching engine across town
Predictable latency matters as much as speed
AWS Outposts brought the same AWS control plane, APIs and security posture into Gemini's NY5 data center cage
51 institutional trading teams were onboarded in a week instead of months

Why it matters

Cloud-native traders have historically accepted a latency penalty as the cost of flexibility. Gemini's architecture closes that gap – delivering sub-2-millisecond access with minutes of setup, versus months of physical infrastructure commitment. For any exchange or financial platform weighing cloud adoption, this shows there no longer needs to be any trade-off between speed
and accessibility.

Read more

Generali – a 194-year-old insurance group serving 71 million customers across 50+ countries – needed to scale AI across 40+ operating entities without duplicating effort or losing consistency. By building a central data, analytics, and AI foundation on AWS, they standardized implementation across claims, underwriting, and operations, and grew their AI application portfolio from 5 to over 50 in three years. The target: 200 applications within the next three.

Generali: Building a central data, analytics and AI solution with AWS

"Using AWS tools and functionality, we can customize the architecture to build the solution we want in terms of features and quality. This keeps costs lower than adopting packaged solutions."
Andrea Pietrasanta, Group Head of Data, AI and Automation, Generali

Key takeaways

€200M+ saved in operating costs in three years across 16 flagship AI use cases
Claim settlement time reduced from several days to one day – and to seconds for simple health claims
21 million shared services processed per month via automated API calls
AI application portfolio grew from 5 to 50+ in three years, with 200 targeted in the next three

Why it matters

Generali's journey shows what becomes possible when AI is built on a common foundation rather than deployed in silos. Standardizing the data and AI layer across 40+ operating entities meant every use case – from pricing to claims to customer service – could be built once and scaled everywhere. For any large, multi-entity financial institution, that architecture decision is what separates isolated pilots from enterprise-wide transformation.