Production Engineering
Your sync layer is part of the product.
July 9, 2026 · 5 min read
Engineers tend to describe synchronization in infrastructure terms: queues, workers, retries. Users describe it in one word — whether they trust the app. That gap is where a lot of reliable systems still feel unreliable.
Sync is visible whether you design it or not
If a record was saved on the device but has not reached the server, the user will find out eventually — usually when a colleague cannot see it. The only choice is whether the product tells them first, calmly, or whether they discover it and draw their own conclusions.
A small vocabulary of states
Most products need no more than six: saved on device, syncing, synced, waiting for connection, needs attention, and conflict. Each should have one consistent visual treatment and one line of plain copy.
Operators are users too
Support and operations teams need their own view: which devices have stuck queues, which mutations failed permanently, which conflicts are waiting. Without it, every sync issue becomes an escalation to engineering, and the sync layer quietly becomes the most expensive feature in the product.
I build and lead teams around systems where reliability matters. If you're designing an offline workflow, evaluating an architecture, or untangling a queue that has become part of your product, I'm happy to compare notes.
More writing
- React Native Architecture Decisions That Become Expensive Later
React Native · 7 min read
- Why Good Engineers Care About Failure Modes
Distributed Systems · 5 min read
- What Engineering Managers Should Actually Measure
Engineering Management · 6 min read