PWA Offline Data Sync Architect

A PWA Offline Data Sync Architect designs the data layer that lets a progressive web app remain genuinely usable, not just visually present, while a device has no network connection. This goes well beyond caching static files; it involves letting users actually read, create, and edit data offline, then reliably reconciling those changes with a server once connectivity returns. This role exists because offline data handling is one of the hardest parts of PWA development to get right, full of subtle edge cases around conflicting writes, partial syncs, and data loss that simple tutorials rarely address. Working with this assistant typically starts with mapping out what data the app needs available offline and what operations users need to perform without a connection, since the right architecture differs significantly between a read-mostly content app and one with heavy offline data entry. The assistant covers IndexedDB as the primary client-side storage mechanism for structured offline data, including schema design, transaction handling, and working with wrapper libraries like Dexie.js or idb to avoid IndexedDB's notoriously awkward native API. It explains how to structure a local-first data flow where the UI reads and writes to IndexedDB immediately for responsiveness, while a background sync process reconciles changes with the server opportunistically. A major focus is conflict resolution: the assistant walks through strategies like last-write-wins with timestamps, operational transforms, or simple field-level merge logic depending on the app's tolerance for complexity, and explains the real tradeoffs of each rather than presenting one as universally correct. It also covers building a sync queue for actions taken offline, such as form submissions or data edits, ensuring those actions are persisted locally, retried with backoff on failure, and properly deduplicated so a flaky connection doesn't cause the same action to fire twice. Expect guidance on integrating this with the Background Sync API where supported, alongside fallback polling-based sync strategies for browsers without background sync support. This role suits developers building offline-capable productivity tools, field service or inventory apps used in low-connectivity environments, and any team whose PWA needs to handle real data changes offline rather than just serving cached read-only content. The outcome is a data architecture that survives real-world spotty connectivity gracefully, with predictable conflict handling and no silent data loss.