One year ago, we introduced AWS Transform for .NET, Mainframe, and VMware workloads—the first agentic AI service designed specifically for large-scale enterprise application modernization. At re:Invent 2025, we rolled out AWS Transform Custom, which lets organizations modernize and transform code at scale using either AWS-managed or custom-built transformations. This includes upgrading language versions, migrating frameworks, boosting performance, and analyzing codebases with transformations that are either ready-made or tailored to your organization’s unique needs. We also added full-stack Windows modernization, Reimagine capabilities, and automated testing for mainframe environments.
In just 12 months, thousands of customers have migrated hundreds of thousands of servers, saved over 1.6 million hours, and processed more than 4.5 billion lines of code using AWS Transform. To mark its first anniversary, AWS Transform agents are now integrated into Kiro, Claude, Cursor, and Codex—including the Kiro-powered agent builder toolkit for creating custom transformation agents.
To discover what’s happened over the past year, the four key lessons we’ve learned, and how those insights shaped our roadmap, check out the one-year anniversary blog post.
Last week’s launches
Here are the announcements from last week that stood out to me:
- Claude Platform on AWS is now generally available – You can now access Anthropic’s native Claude Platform experience—including APIs, console, and early-access beta features—directly through your existing AWS account. No need to manage separate accounts, billing, or usage tracking. Note: Claude Platform on AWS is operated by Anthropic, and customer data is processed outside the AWS security boundary. For details, read the deep dive blog post.
- Amazon EC2 M3 Ultra Mac instances – Powered by Apple M3 Ultra Mac Studio hardware, these instances feature a 28-core CPU, 60-core GPU, 32-core Neural Engine, and 256GB of unified memory. Compared to EC2 M4 Max Mac instances, M3 Ultra offers double the unified memory, 1.75x more CPU cores, 1.5x more GPU cores, and twice the Neural Engine cores—giving Apple developers the resources to run many more Xcode simulators at once and speed up on-device ML workflows, helping bring products to market faster.
- Amazon Redshift RG instances with AWS Graviton – These instances deliver up to 2.4x better performance for data warehouse and data lake workloads compared to previous-gen RA3 instances, at a 30% lower cost per vCPU. RG instances include Redshift’s custom vectorized engine for querying Apache Iceberg and Parquet data directly on cluster nodes.
- Amazon Bedrock Advanced Prompt Optimization – Optimize your prompts for any model on Bedrock, and compare your original prompts against optimized versions across up to five models at the same time. Ideal if you’re switching models or simply want better results from your current one.
- AWS Security Agent full repository code scanning (preview) – A new feature in AWS Security Agent performs deep, context-aware security analysis across your entire codebase. When it finds vulnerabilities, it generates precise code fixes tied to specific files and lines—helping teams resolve security issues faster than ever. This capability is free for existing AWS Security Agent customers during the preview period.
- AWS Interconnect – multicloud connectivity with Oracle Cloud Infrastructure (preview) – Quickly set up resilient, scalable private connections to other cloud providers using AWS Interconnect. Oracle Cloud Infrastructure (OCI) is the latest cloud provider to adopt the open specification behind AWS Interconnect. This gives AWS customers a consistent, simple experience when connecting to OCI (preview), Google Cloud (generally available), and Microsoft Azure (expected later in 2026).
Additional updates
Here are a few more news items you might find useful:
- Accelerate AI research and education with the Build on Trainium program – See how the next generation of AI researchers is leveraging Amazon’s custom chips to speed up innovation. AWS has invested $110 million to provide university researchers with access to purpose-built AI hardware. AWS Trainium is already accelerating AI research at institutions like UC Berkeley, MIT, and Carnegie Mellon. All findings are open source, so improvements benefit the entire developer community.
- Complete list of AWS Community Days 2026 – There’s a special energy at events where speakers are fellow practitioners, organizers are passionate volunteers, and the agenda is shaped by the community. That’s the spirit of AWS Community Days—and they’re taking place in cities on every continent this year.
- The Kiro Startups Credit program returns – After thousands of founders applied in the first round, applications are now open again. Apply to receive up to one year of Kiro Pro+ credits automatically applied to your organization’s AWS account.
For a full roundup of recent AWS blog posts, be sure to visit the AWS Blogs page regularly.
Explore more about AWS, browse upcoming in-person and virtual events—including AWS Summits, startup gatherings, and developer-focused sessions—and join the AWS Builder Center to connect with fellow builders, share solutions, and access resources to support your development journey.
That wraps up this week’s update. Tune in next Monday for another Weekly Roundup!
— Channy
