Prerequisites

Prepare the server before you enable Easy8 AI.

Base requirements

• Linux server with Docker Engine and the Docker Compose plugin (docker compose)
• persistent Docker storage for service data
• Nginx or another reverse proxy in front of Easy8
• enough CPU, RAM and disk space for the product you install

Network requirements

The server must allow:

• browser access to Easy8 over HTTPS
• reverse proxy access to the chain service on the host
• internal communication between app, chain, celery, redis and, for PRO products, chroma

For cloud AI, the server also needs outbound access to OpenAI or Azure OpenAI.

For self-hosted AI, the chain and celery containers must reach the LLM endpoint (LLM_URL) and, for PRO products, the embeddings endpoint (EMBEDDINGS_URL).

Services by product

Product	Required services
Easy8 AI Basic	chain, celery, LLM provider
Easy8 AI Knowledge assistant PRO, Easy8 AI Helpdesk	chain, celery, chroma, embeddings provider, LLM provider

Easy8 AI Basic does not use chroma or embeddings.

Self-hosted AI

If you run the LLM or embeddings yourself, prepare:

• a server with an NVIDIA GPU and enough GPU memory for the model
• NVIDIA drivers
• NVIDIA Container Toolkit

Embeddings (BAAI/bge-m3 served through TEI) run on NVIDIA GPU only; CPU-only deployment is not supported. See Embeddings hardware recommendations and LLM hardware recommendations, then continue with Self-hosted AI server setup.

Then pick your guide in Choose a scenario.