Fix AI App Builder Performance

My cross-platform AI-driven mobile app is suffering from performance issues that trace back to the data layer. On both iOS and Android, new records intermittently fail to save, and previously stored data sometimes disappears or reappears only after a forced refresh. This “database and saving” problem is the root cause of user-visible lag and occasional freezes when the app attempts to read or write.

What I need:
• Pinpoint why data is not saving correctly in the current database implementation and resolve it without breaking existing features.
• Optimise the read/write cycle so that loading screens and UI interactions stay smooth even under poor network conditions.
• Ensure fixes work consistently on the latest production versions of iOS and Android, then regression-test earlier OS versions we still support.
• Provide a concise technical summary of the root cause, the changes made, and any future maintenance recommendations.

Acceptance criteria:
1. New and existing records must save and sync reliably on both platforms after multiple stress-test cycles.
2. Average write and retrieval times should drop to an acceptable threshold agreed during kickoff (benchmarked with real-world sample data).
3. No new crashes or memory leaks introduced, verified with Xcode Instruments / Android Profiler (or equivalent).
4. Final build passes store-submission automatic checks for iOS App Store and Google Play.

Code access is ready in Git, plus a small test suite you can extend. Most of the stack is documented; any gaps can be clarified quickly. Once the fix is in place, I’ll take care of redeploying to production—just need a stable build and your implementation notes.