AWS Transform Reaches One-Year Milestone with Record Adoption
AWS Transform, the agentic AI service for modernizing enterprise applications, has processed over 4.5 billion lines of code and saved customers more than 1.6 million hours in its first year, the company announced today.
The service now integrates with Anthropic's Claude, Kiro, Cursor, and Codex, enabling developers to build custom transformation agents directly within their preferred AI environments.
“In 12 months, we've seen thousands of customers migrate hundreds of thousands of servers. The speed and scale at which enterprises are modernizing is unprecedented,” said Dr. Sarah Chen, VP of Modernization Services at AWS. “The integration with Claude and other agent platforms marks a new chapter in automated code transformation.”
Background: What AWS Transform Offers
Launched at re:Invent 2025, AWS Transform is an agentic AI service designed for automating enterprise application modernization. It supports .NET, Mainframe, and VMware workloads, and later added AWS Transform Custom for organizations to create and apply custom code transformations.
Key capabilities include migrating framework versions, optimizing performance, and enabling full-stack Windows modernization. The service also provides automated testing for mainframe migrations.
What This Means for Enterprise Modernization
The combination of AI agents and AWS Transform allows companies to modernize legacy code at scale without manual rewrites. With integrations into Claude, Kiro, Cursor, and Codex, developers can now trigger transformation workflows directly from their AI tools.
“This reduces the friction between ideation and execution. We expect to see even faster adoption across regulated industries like finance and healthcare,” Chen added.
Claude Platform on AWS Now Generally Available
Anthropic's native Claude Platform experience is now directly accessible through AWS accounts, eliminating the need for separate billing or management.
The platform includes full access to Claude's API, console, and early-access beta features. Data is processed outside AWS's security boundary, with Anthropic operating the service.
Amazon EC2 M3 Ultra Mac Instances Launched
New EC2 M3 Ultra Mac instances, powered by Apple M3 Ultra chips, offer 2x unified memory (256GB), 1.75x more CPU cores (28-core), and 2x Neural Engine cores compared to previous M4 Max instances.
These instances enable developers to run significantly more Xcode simulators in parallel and accelerate on-device machine learning workflows, reducing time to market for Apple applications.
“Apple developers can now push the boundaries of what's possible with unified memory and GPU performance,” said Mark Larson, Principal Engineer at AWS. “M3 Ultra instances are a game-changer for iOS and macOS app testing.”
Amazon Redshift RG Instances with Graviton Deliver 2.4x Performance
Redshift RG instances, powered by AWS Graviton processors, run data warehouse and data lake workloads up to 2.4x faster than previous RA3 instances, at 30% lower cost per vCPU.
The instances include a custom vectorized query engine for Apache Iceberg and Parquet data, optimizing performance for analytical workloads.
Bedrock Advanced Prompt Optimization Launched
Amazon Bedrock now supports Advanced Prompt Optimization, allowing users to optimize prompts for any model on the platform.
Users can compare original and optimized prompts across up to five models simultaneously, facilitating model migration or performance improvements.
Upcoming Coverage
For deeper insights into AWS Transform's roadmap and lessons learned, read the one-year anniversary blog post. For a detailed breakdown of the Claude Platform integration, see the deep dive.