There's a "saying" that with the migration to the cloud we've started treating our servers as cattle instead of pets, and now with the migration to serverless we're treating them like cockroaches :)

The thing is... infrastructure costs money, and paying for infrastructure when you're not using it is wasting money. This brings us to the conclusion that there's no sense in having pets around the world if we only need them once in a while. The problem is that if we scale down our VM's and a considerable load comes, we won't be able to scale up quickly or even scale up at all if the process is only manually triggered. That's precisely the topic of this talk :)

We're gonna talk about ideas and concepts on how to run a system in a way that we provision, deploy, and scale automatically, in time and without any stress. You'll see that being in a single region or all over the world, on one cloud or on a bunch of clouds, is but a small detail. Yeah, you're gonna hear about containers, serverless, managed services, event-driven, etc... but don't worry. We're going to focus on the abstract idea and concepts of how that system works, communicates, fails, heals, etc., and we can discuss the tech details afterward ;)