FF News — The Fintech News Network

ANZ Accelerates Cloud Transformation with AWS Using Agentic AI to Migrate 1.7 Petabytes of Data

By Lauren Towner · 19 November 2026

In brief: ANZ has migrated 1.7 petabytes of data to AWS in just six months using agentic AI, reducing migration effort by 35% and cutting daily operating costs by over 70%.

Key facts
Data Volume1.7 petabytes
Migration Timeline6 months
Cost ReductionFrom $10,000/day to <$3,000/month
Efficiency Gain35% reduction in migration effort
Press Release: ANZ Accelerates Cloud Transformation with AWS Using Agentic AI to Migrate 1.7 Petabytes of Data | Featured Image by FF News

ANZ has successfully migrated 1.7 petabytes of data to Amazon Web Services (AWS), completing a massive transition of its Enterprise Big Data platform in just six months. By utilizing agentic AI to automate complex engineering tasks, the bank reduced its expected migration timeline by a year and slashed daily operating costs from $10,000 to less than $3,000.

What was announced

The migration involved moving ANZ’s Enterprise Big Data platform—which underpins critical services for payments, business banking, and customer remediation—from an on-premises environment to Amazon EMR. The scale of the project was significant, encompassing 300 million files, 142,000 tables, 4,500 automated data workflows, and 4,000 views. The 1.7 petabytes of data transferred is roughly equivalent to more than 340 million smartphone photos.

A central component of the project was the use of AI-assisted engineering to accelerate delivery. Engineers employed a troubleshooting assistant powered by Amazon Bedrock to identify and resolve performance bottlenecks, eliminating the need for manual code reviews. This application of agentic AI across discovery and migration activities reduced the total migration effort by an estimated 35 per cent and halved the number of defects and rework through automated testing and validation.

The new cloud-based architecture has already demonstrated performance gains, processing data approximately seven times faster than initial alternative methods. To date, the platform has tokenized nearly 80 terabytes of data. Beyond the immediate technical improvements, the transition has moved the platform’s operating costs from more than US$10,000 per day to less than US$3,000 per month, providing a more sustainable financial model for the bank’s data infrastructure.

"Data is the foundation of modern banking and increasingly the foundation of AI. As we explore new ways to use data and analytics to improve customer and employee experiences, having modern, scalable and secure infrastructure is critical."

Kai Yang, Chief Data and AI Officer at ANZ

The companies involved

ANZ, or the Australia and New Zealand Banking Group, is one of the "Big Four" Australian banks and a major financial institution in the Asia-Pacific region. The bank is currently executing its 2030 Strategy, which focuses on simplifying its technology estate and modernizing critical infrastructure to support large-scale AI and data initiatives. ANZ has established itself as a leader in digital experimentation, particularly in the realms of tokenization and cloud-native banking services.

Amazon Web Services (AWS), a subsidiary of Amazon, is the world’s most broadly adopted cloud platform. It provides over 200 fully featured services from data centers globally, including the Amazon EMR and Amazon Bedrock tools used in this migration. AWS Professional Services worked directly with ANZ engineers to implement the AI-assisted migration framework. The company maintains a dominant position in the financial services cloud market, frequently partnering with Tier 1 banks to manage the transition from legacy on-premises systems to scalable cloud environments.

What FF News has reported before

FF News has closely tracked ANZ’s digital transformation and its ongoing relationship with AWS infrastructure. Recently, ANZ and Swift Complete Landmark Live Tokenised Deposit Transaction for BHP Treasury, demonstrating the bank's appetite for blockchain-based financial innovation. We have also seen AWS deepen its role in the banking sector’s migration efforts; for instance, Thought Machine and AWS Launch AI Tool to Accelerate Legacy Core Banking Migrations, which highlights a broader industry trend of using machine learning to solve the "legacy bottleneck." Additionally, the availability of financial infrastructure on the cloud was further evidenced when OpenAssets Launches Institutional Tokenization Infrastructure on AWS Marketplace.

What this means

The speed of this migration—six months instead of 18—is a significant moment for the industry. For years, the primary barrier to digital transformation for Tier 1 banks has been the sheer "gravity" of legacy data; moving petabytes of information while maintaining real-time service was considered a multi-year, high-risk endeavor. By proving that agentic AI can handle the "heavy lifting" of code conversion and troubleshooting, ANZ and AWS have effectively commoditized the migration process. This puts immense pressure on other global incumbents who are still tethered to expensive on-premises data centers. The massive 70% reduction in daily operating costs further changes the ROI calculation for cloud adoption, moving it from a strategic necessity to an immediate fiscal imperative.

Frequently asked questions

How did ANZ migrate 1.7 petabytes of data so quickly?

ANZ completed the migration in just six months by using AI-assisted engineering and agentic AI. By leveraging an AI assistant powered by Amazon Bedrock, engineers were able to identify and fix performance bottlenecks automatically, reducing the total migration effort by 35% and finishing the project a year ahead of the original 18-month estimate.

What are the cost savings of ANZ's AWS migration?

The migration to AWS has resulted in a massive reduction in operating costs for ANZ's data platform. Expenses dropped from more than US$10,000 per day on the previous on-premises system to less than US$3,000 per month on the new cloud-based architecture.

What is agentic AI in the context of banking?

In this context, agentic AI refers to autonomous or semi-autonomous AI tools used during discovery and migration activities. ANZ used these tools to analyze, convert, and test workloads, which helped reduce defects and rework by approximately 50% while processing data seven times faster than traditional methods.

More AI in Finance news →

About the companies

ANZ

Website · 10 stories and 1 video interview on FF News since 2023

Previously on FF News:

Amazon Web Services

Part of Amazon · Website · 127 stories and 4 video interviews on FF News since 2021

Previously on FF News:

Companies in this story: ANZ, Amazon Web Services

People in this story: Chris Casey, Kai Yang, Donald Patra

More from News