AWS Transform adds field-level tracing for mainframe modernization

A more precise catalog and a coverage report

On 3 November, AWS announced field-level business-function discovery and a requirements coverage report for the assess and reimagine workflow. Previously, functions using the same table or file were grouped together. AWS says the new grouping depends on the data fields affected.

The update includes screens and queues in the discovered functions and links generated requirements to business rules and those rules to source lines. AWS describes line-level classification of how source was handled. The workflow supports modernization assessment for z/OS COBOL and PL/I workloads in regions where AWS Transform for mainframe is available.

This is a product capability announcement. It does not establish that all extracted requirements are correct, or that every resulting modernization preserves the original application’s behavior. Source: AWS Transform improves the mainframe reimagine workflow

Analysis: coverage is useful, but it is not equivalence

A coverage report can answer whether something was considered. It cannot automatically answer whether it was understood correctly. A source line may be linked to a requirement whose wording misses an exception, or a rule may be correctly extracted but implemented differently later. Teams need to keep those questions separate.

An illustrative billing system shows the distinction. Two functions might touch the same customer record while serving different purposes: one changes a billing address and another applies a credit restriction. Grouping them simply because they share a table could obscure the boundary that a redesign needs to preserve. More precise field-level information could help reviewers ask better questions.

However, field separation alone does not prove business independence. A workflow can depend on timing, transaction order or an external process even when individual fields appear separate. Reviewers should use the generated catalog as a starting point and check it against how the organization operates.

The strongest use of traceability is to make disagreement actionable. If a domain expert says a requirement is wrong, the team should be able to follow it back to the relevant code and associated rules, then record the correction. That is more valuable than a coverage percentage that cannot explain why an item was counted.

A focused acceptance test for modernization teams

Start with a bounded area that has reliable expected outputs and access to a knowledgeable owner. Generate the catalog and requirements, then inspect a sample chosen for difficult behavior rather than only common paths. Include an exception case, a queue-driven interaction and a rule whose meaning depends on earlier processing.

Track corrections by type. Missing behavior, misleading grouping and ambiguous language create different downstream work. A single total error count would make it harder to decide whether the tool is useful for assessment, planning or implementation. Measure how quickly a reviewer can navigate from a disputed requirement to its evidence.

Before approving replacement code, retain tests of the original behavior and specify which differences are intentional. Otherwise a rewrite can be declared successful merely because it matches a generated document, even if that document was the faulty step.

A further check is to ask a reviewer unfamiliar with the codebase to follow one requirement back to its evidence. If that person cannot reconstruct the connection, the report may be technically comprehensive but difficult to use. Record where interpretation required an experienced maintainer. Those moments reveal which knowledge remains in people’s heads and should be captured explicitly before the project proceeds to replacement design and implementation.

The announcement addresses the bridge between legacy code and a redesign proposal. Faster requirement generation is helpful; a reviewable explanation of what was retained, changed or omitted is the stronger result to pursue. The commercial value depends on whether that bridge reduces uncertainty for the people accountable for the migration.