HM.
Back to work

JCR-to-Native-AppServlet

Client
Adobe collaboration
Engagement
Current, ongoing

Architected a backend servlet that exports JCR content as structured JSON to feed an existing native mobile app, guided end to end with AI-assisted development.

  • Backend architecture
  • JCR to JSON export
  • AI-agent-guided development
  • Native app integration
1:1
web block to native component mapping
0
code changes needed for a new component type

This backend servlet reads JCR content from an Edge Delivery Services site, authored through Adobe's Universal Editor, and exports it as structured JSON. That JSON feeds an existing native mobile app, mapping web block structure directly onto native app components: a hero block becomes a native banner component, and so on.

The system is designed for extensibility. Supporting a new component type only requires adding a JSON mapping, with no code change required. It was built in collaboration with a senior full-stack developer for backend implementation depth; the core servlet logic is complete and currently being validated against each migrated site.

Worth naming directly: this was delivered through architecture ownership and AI-agent-guided implementation, not hand-written Java. It is a legitimate and interesting story about directing AI-assisted development toward a production backend service, not something to obscure.

What shipped

  • Reads JCR content authored via Adobe's Universal Editor
  • Exports structured JSON consumed directly by a native mobile app
  • New component types are supported by mapping, not by code changes

Edge Delivery site

Content authored via Universal Editor

JCR servlet

Reads JCR content, maps blocks to components

Structured JSON

One mapping file per component type

Native mobile app

Renders native components from the feed