Document how to stand up new Animl instances

[nathanielrindlaub]

We don't maintain great documentation on how to deploy fresh instances of Animl to AWS for users who are interested in standing up their own infrastructure. Most of the infra is managed in Serverless config files, but some definitely is not. For example, off the top of my head:

• MongoDB Atlas (and soon a new Atlas Search index)
• Cognito user pool (currently the CF template just lives in our AWS account and is not in version control)
• All Sagemaker models, endpoints, and corresponding SSM params
• SES config (?)
• Route53 config (?)

Additionally, some of the Serverless-managed stacks are only necessary if users need to consume wireless images (e.g. Animl Email Relay and the programmatic cell cam ingestion api)

I think we should step through the process of standing up a separate instance and thoroughly document the steps to set up all of these external pieces, and do what we can to internalize them in some kind of infrastructure as code solution going forward.

Related: #58

[nathanielrindlaub]

@lessej pointed out it would be great to have a diagram outlining how all the micro-services fit together too.