AI-Enabled Legacy Modernisation Playbook

Version 1.0.0

This playbook describes how Defra’s Legacy Application Programme (LAP) uses AI to reverse-engineer legacy applications and rebuild them. It covers the end-to-end process — from gathering source material through automated analysis to a signed-off PRD, feature specifications, and an autonomous build loop that implements the replacement application feature by feature.

flowchart LR
    A["Gather Inputs"] --> B["Content Curation"]
    B --> C["Review Curated Outputs"]
    C --> D["Analysis & PRD Generation"]
    D --> E["PRD Review & Sign-off"]
    E --> F["Feature Decomposition"]
    F --> G["Rebuild Setup"]
    G --> H["Autonomous Build"]

Sections

  • Overview — what this playbook covers, team structure, and stakeholder roles
  • Process — step-by-step guide through each phase of the reverse engineering process
  • Tooling — AI coding assistants, plugins, and project directory structure
  • Output Reference — detailed descriptions of each artefact the process produces
  • Considerations & Caveats — information governance, PII handling, AI quality, and costs
  • Glossary — key terms and definitions
  • Contributing — how to propose changes to this playbook

Start with the Overview for context, then follow the Process guide.