
Designing Idempotent Background Workers for LLM Calls in Rails
An LLM call behind a Sidekiq job runs at least once, not exactly once. How to make Rails workers idempotent so a retry never double-classifies or double-bills.
Tag archive

An LLM call behind a Sidekiq job runs at least once, not exactly once. How to make Rails workers idempotent so a retry never double-classifies or double-bills.
Some slowness shouldn't happen in the request at all. Moving slow work (email, third-party APIs, reports) to background jobs — plus the traps: idempotency, passing ids vs objects (GlobalID), queue priorities, retries.

You've built a checkout reactor that reserves inventory, charges a card, generates a shipping label,...

Ruby Reactor 0.4.0 introduces interrupts — the first native saga pause/resume mechanism in...

🚀 Ruby on Rails Makes Building Products Feel Fast & Powerful One thing that continues to stand...

In modern web applications, keeping the Request-Response cycle fast is crucial for user experience....

It’s been a minute since I last published a blog post, mostly because we’ve been busy building...

In the Rails ecosystem, job processing libraries are essential for managing background tasks...
Run Sidekiq 6 as daemon in Production environment on Ubuntu...

TL;DR: In this post I explain how you can have persistent connections to Redis using Async::Redis...
Hey there! Misha Merkushin here, Team Lead of the Ruby Platform team at Kuper Tech. We’re the crew...
Background processing is a crucial part of modern web applications. Tasks like sending emails,...