AWS Transform
AWS Transform for Java: what it does and where it needs engineers.
It rewrites code and reports what it changed. It cannot prove the result is right for your system. That part is ours.
The short answer
AWS Transform is an AWS service that uses AI to automate code modernization tasks, including Java version upgrades. It rewrites code and reports what it changed. It cannot prove the result is correct for your system. Ace8 runs it against your tests and has senior engineers review every change.
Definition
What AWS Transform is
AWS Transform is an AWS service for automating modernization work with AI. For Java teams, the most common use is upgrading the Java version of an application. You point it at a codebase, it proposes changes, and it summarizes what it did. It is also referred to by an earlier name, Amazon Q Code Transformation.
Ace8 is not affiliated with AWS. This page describes how we run the tool on client code, not a partnership. We have no commercial interest in whether you pick it.
Scope
What it can transform
Check AWS documentation for the current list of supported source and target versions, because it changes. In practice, the tasks suited to this kind of tool are the mechanical ones.
- Java version upgrades across a codebase.
- Replacing APIs removed or deprecated in newer Java releases.
- Dependency version updates tied to the upgrade.
- A change summary the reviewer can read before opening the diff.
Our process
How we run it on a client codebase
- 1
Choose the slice
One application or module at a time, small enough for a person to review the whole diff.
- 2
Baseline the tests
Run the existing suite and record what passes. Where coverage is thin, write characterization tests first.
- 3
Run the tool
Take its proposed change on a branch, never directly on the main line.
- 4
Build and test
Compile, run the suite and compare with the baseline. Anything that differs is a finding.
- 5
Engineer review
A senior engineer reads the diff and the tool's summary, fixes what is wrong by hand and signs the result.
Evidence
Results on our sample repo
We are running AWS Transform on the same open-source Java 8 and Spring Boot 2 application that we use for OpenRewrite and GitHub Copilot app modernization. We will log files changed, test results before and after, manual fixes and engineer review minutes, and publish them here. Until then we do not claim a result. See the AI-assisted overview.
Limits
What a tool like this misses, and why
These are failure classes that any automated upgrade can hit. We list what to look for, not what it did on your code.
- Behavior that lives in runtime configuration rather than code.
- Changes that compile but alter behavior, such as default settings in a newer library.
- Libraries with no upgrade path that the tool can follow.
- Test code and build scripts, which are often treated differently from production code.
- Anything not covered by tests, where neither the tool nor the reviewer has a signal.
Choosing
When OpenRewrite or Copilot fits better
Read about OpenRewrite next. Many projects use both.
| Situation | Better fit | Why |
|---|---|---|
| You need a repeatable, auditable result across many repositories | OpenRewrite | Recipes are deterministic: the same input gives the same output |
| A specific migration has a published recipe, such as Spring Boot 3 | OpenRewrite | The recipe is the migration |
| The codebase mixes patterns a recipe will not match | AWS Transform or an assistant | Flexible, but its output needs closer review |
| Developers want help inside the editor on one project | GitHub Copilot app modernization | Fits the individual workflow |
Cost
Cost and effort factors
Check current AWS pricing for the service itself. The larger cost is usually engineer time, and four things drive it.
- Test coverage, because weak tests mean slow, uncertain review.
- Codebase size and how uniform it is.
- The number of third-party libraries that need separate handling.
- How far the source version is from the target.
It is an AWS service that uses AI to automate code modernization tasks, including Java version upgrades. It proposes code changes and reports what it changed.
No. It cannot prove its output is correct for your system. You need tests that pin down current behavior and an engineer who reviews the result.
No. Ace8 is not affiliated with AWS. We run the tool on client code as one of several options.
Yes. The assessment includes a trial on a representative sample, with the results written up.
OpenRewrite uses deterministic recipes and gives the same result every run. AWS Transform is AI-driven and more flexible, with output that needs closer review. We often use both.
Know the cost before anything moves.
A fixed-price assessment gives you the inventory, the risk score and a phased plan. Scope and cost are fixed before the migration starts.
Book a Fixed-Price AssessmentGet in touch
Talk to a modernization engineer
Tell us what you run today. An engineer, not a sales script, replies within one business day.