AI-Assisted Custom Code Migration to SAP S/4HANA: ATC Findings, Quick Fixes and the Custom Code Migration Agent

Keep this: job aid

AI-assisted custom code remediation: run card

• Scope first: SCMON and SUSG usage for a year; delete unused code with the app's deletion transport.
• Analyze with S4HANA_READINESS_2025 (or _NO_FLE) from the central check system; import the Simplification Database in SYCM.
• Prioritize: priority 1 and 2, then usage, quick-fix availability, clean core level and criticality.
• Use Joule Explain to understand findings; verify explanations against code and process owners.
• Apply deterministic quick fixes first (Recommended Quick Fixes, quick fix comments on), then recheck.
• AI proposals for the long tail only; review each line; high-risk code stays with its owner.
• Agent prerequisites: S/4HANA 2023+ target, ADT for Eclipse with MCP, Joule for Developers, Copilot Pro or Amazon Q Developer Pro, SAP Notes 3751537 / 3751829 / 3759994.
• One package, one dedicated request, MCP access restricted to the package; run in sandbox or CMD only.
• Evidence: unified diff reviewed, comments resolved, before-and-after comparisons with explained differences, characterization tests.
• Measure: throughput with review, auto-applied and comment shares, rework, escaped defects, review hours, AI Units.