Could this be done using configuration management (CM) scripts, like Saltstack, Ansible or Puppet?
If you don’t have infrastructure as such, surely you need configurations for various micro or nano services?
Just thinking out lound..
As cool as start-ups are, I’d urge caution and suggest anybody building services (apps) on SaaS or PaaS pay close attention to disaster recovery options in terms of how their business continues in absence of a critical service like AWS Lambda or whatever (Slack, …).
So other people’s servers then? :-p
I mean, AWS Lambda is made up of servers (many!) and presented to users as a service where they don’t have to worry about config and maintenance.