What is this?

Your entire AI stack, containerised and production-ready.

Docker containers package your AI services into isolated, portable units that run identically every time. Instead of installing software directly on macOS where updates and conflicts can break things, each service (Open WebUI, n8n, databases, vector stores) runs in its own container with its own dependencies — completely isolated from your system.

With HKMAC AI's Docker Stack setup, we configure a complete docker-compose.yml file that defines your entire AI infrastructure as code. One command brings everything up. One command takes it all down. Need to move to a new Mac? Copy the compose file and you are running in minutes, not hours.

Docker on Apple Silicon runs natively through Docker Desktop with ARM64 containers, giving you near-native performance. We optimise resource allocation, volume mounts, and networking to ensure your containerised AI services perform as well as native installations while being far easier to maintain, back up, and replicate.

This is especially valuable for teams running multiple AI services. Instead of troubleshooting Python version conflicts and dependency issues, you get a clean, declarative infrastructure that just works. Update a service? Change one line. Roll back? One command. It is DevOps best practice, applied to your local AI setup.

How it works

One compose file. Entire AI stack.

graph TB A["📄 docker-compose.yml"] --> B["🐳 Docker Engine"] B --> C["📦 Open WebUI
Container"] B --> D["📦 n8n
Container"] B --> E["📦 ChromaDB
Container"] B --> F["📦 PostgreSQL
Container"] C --> G["🌐 Shared Network"] D --> G E --> G F --> G

What you get