Docs: Add Documentation for Deploying with ai-accelerator

[aslater-rh]

What does this PR do?

This is a doc-only PR containing a guide on how to deploy the ai-accelerator with the Layered Zero Trust Validated Pattern. More than happy to create an accompanying issue for this - let me know if that is desired for traceability.

What I've Tested

I've followed the steps here (and in the accompanying ai-accelerator PR multiple times on the demo platform and was able to verify the following:

1. LZTVP deploys with no errors - both while running the script and after the operators are installed (no Argo sync errors). I was testing off a older version of LZTVP main so I am leaving this in draft until I re-verify with the most updated versions of both repos.
2. All LZTVP applications can fully synchronize.
3. No failing pods or pods stuck in initializing.
4. Once you disable the openshift-gitops-operator Argo application on the ai-accelerator side there are no Argo synchronization errors (with the exception of the ai-accelerator-operator which tries to synchronize indefinetely on the demo platform but not in IBM Cloud which is where we're deploying this for our customer)

We're also having a discussion about Zero Trust as it pertains to model serving later this afternoon - will be happy to add some outputs of that conversation here if that is also desired.