Learn Cadence Workflow (https://cadenceworkflow.io/). It implements a new way to build distributed systems using fault tolerant virtual memory. It lowers the complexity bar for hardcore backend programming by eliminating the need for managing the application state through databases and queues.
This has been so useful for us. This gives us the ability to write synchronous code that can literally fall asleep for days, wake back up and run some more stuff, fall back asleep, and be totally scalable. It changes the game.
I mean systems that have to scale to hundred of millions of users and guarantee durability in presence of infrastructure and software failures.
Cadence is used in production by over 100 teams at Uber and multiple outside companies. For example here is a presentation that describes how HashiCorp relies on Cadence for the new version of its Consul product: https://www.hashicorp.com/resources/making-multi-environment.... The Cadence related part starts at 17:30.