This talk is about solving real-life scaling problems! I share my experience of our systems hitting performance issues with Sidekiq queues and complex validations. Simply adding batching, breaking jobs into chunks, or fine-tuning retry strategies overloaded the Rails app — and didn’t help.

We evaluated multiple options, but found that decoupling the heavy tasks into Go allowed us to keep Rails nimble while boosting performance.

I discuss how we use Go with Rails to handle jobs that are compute-heavy to harness the raw speed, concurrency and low memory instead of tapping into ActiveRecord or Rails’ magic.

Key takeaways:
* Gracefully evolve Rails to not do heavy lifting!
* Learn Patterns to offload CPU-bound or IO-bound tasks using Go
* Managing communication between heterogeneous tech stacks.