Is there a recommended system setup.

[Jyrbian]

Any recommendations for setting up a Dev environment for BMAD?

After finding the BMAD method; I have been wondering if I am overthinking my setup and wondering how others are setting up their development environment.

I have been building a local AI and development cluster for agentic coding using Kubernetes. I am having some difficultly getting it setup due to version incompatibility with GPU pass through for ROCm either in the docker image or through Kubernetes. I have gotten it to work with ComfyUI and Ollama, so I will eventually get it all setup.

[My planned setup]
I am using a local first approach, with access to cloud AI.

Docker images on Kubernetes cluster:
Open WebUI: Talk to local AI models, use API form ComfyUI to generate pics.
LiteLLM: An AI proxy, that routes requests to the AI service (local with vLLM or Ollama, or cloud like Open AI, Claude, etc..)
Model Server: vLLM, TGI, Ollama, etc, a local model server that the other
Qdrant: The Vector Database for long-term memory for your AI, and Retrieval-Augmented Generation.
LangChain Orchestrator: determine if a request needs to talk to the model server, search the Qdrant database, or use another tool. I might try volcano here as an alternative.
ComfyUI: The Image Generator.
Dev-Env: A dedicated container for agentic coding.

[bmadcode]

Might get more response in the discord - just reopened the forums here a day or two ago - might be a bit slow getting traction here again if looking for a faster response.