How to Deploy EdgeQuake with Docker: Step-by-Step Tutorial

Deploying a GraphRAG stack sounds intimidating — a Rust backend, a frontend, a database, vector indexes... but EdgeQuake was built to remove all of that friction. In this tutorial, you'll have a full GraphRAG pipeline running in about five minutes using nothing but Docker.

Prerequisites

• Docker + Docker Compose (v2)
• An OpenAI API key, or a local Ollama instance with an LLM + embedding model
• 2+ CPU cores and 4 GB RAM minimum (4 cores / 8 GB recommended)

Step 1 — One-line quickstart

EdgeQuake ships an interactive installer that detects your environment, asks about your LLM provider, and starts the whole stack:

curl -fsSL https://raw.githubusercontent.com/raphaelmansuy/edgequake/edgequake-main/quickstart.sh | sh

Prefer manual control? Grab the compose file directly:

curl -fsSL https://raw.githubusercontent.com/raphaelmansuy/edgequake/edgequake-main/docker-compose.quickstart.yml -o docker-compose.quickstart.yml
docker compose -f docker-compose.quickstart.yml up -d

Step 2 — Headless / CI setup

No interactive terminal? Set the provider via environment variables:

# OpenAI
EDGEQUAKE_LLM_PROVIDER=openai OPENAI_API_KEY=sk-... docker compose -f docker-compose.quickstart.yml up -d

# Ollama on host
EDGEQUAKE_LLM_PROVIDER=ollama EDGEQUAKE_LLM_MODEL=gemma4:e4b \
EDGEQUAKE_EMBEDDING_PROVIDER=ollama OLLAMA_EMBEDDING_MODEL=embeddinggemma \
docker compose -f docker-compose.quickstart.yml up -d

Step 3 — Verify and start ingesting

Check the API health endpoint, then open the web UI:

curl -s http://localhost:8080/health | python3 -m json.tool

In quickstart mode the API is open (no login) — perfect for local experiments. Upload PDFs, let EdgeQuake build the knowledge graph, then ask multi-hop questions that plain vector search would miss.

Production notes

That's it — a complete, self-hosted GraphRAG stack in minutes. For a pinned version, run EDGEQUAKE_VERSION=0.26.0 sh quickstart.sh. Happy graph-building! 🌋