Adding a Fallback Model to Hermes with Amazon Bedrock

I already have Hermes running against a model I configured on Azure AI Foundry, and to be fair it works. The catch is in the word "already": one provider means one point of failure. A quota cap, a throttled endpoint, or a bad afternoon on someone else's cloud, and the assistant I lean on all day just sits there returning errors. Leave the Azure AI Foundry setup exactly as it is, then wire Amazon Bedrock in behind it as a fallback, so Hermes has somewhere else to go when the primary provider is not answering