Deployment checklists

Local development

Prerequisites installed Environment configured Database started API ready Frontend running Backend reachable Tests passing

Apple Silicon

Native Ollama installed Authorized model installed Acceleration checked where observable Model API responding Backend connected Visible streaming verified

NVIDIA — unverified target

GPU detected Driver qualified Container GPU access verified vLLM started Model loaded Inference verified Memory measured API connected

Production — qualification pending

Secrets configured TLS qualified Migrations applied Access controls tested Monitoring enabled Backup restored in isolation Recovery reviewed